energy e ciency improvement in coal fired power plant ...certi cation of dissertation i certify that...
TRANSCRIPT
University of Southern Queensland
Faculty of Health, Engineering & Sciences
Energy Efficiency Improvement in Coal Fired Power Plant
through Operational Optimisation
A dissertation submitted by
Jason Lang
in fulfilment of the requirements of
ENG4112 Research Project
towards the degree of
Bachelor of Electrical Electronic Engineering
Submitted: October, 2015
Abstract
Energy consumption is a significant cost to all business with the countries large industrial
plants consuming 75 % of all energy produced in Australia. This cost is not only a finan-
cial burden but it has an environmental cost. The energy consumption within coal-fired
power stations that is directly associated with generation is called auxiliary power. Ap-
proximately 10 % of all power produced is used to drive power stations internal auxiliary
power needs. It is the auxiliary power consumption at Tarong Power Station that is the
focus of this dissertation.
This dissertation first seeks to understand the stations energy consumption through a
comprehensive review of auxiliary power issues worldwide and the creation of control
system tracking logic. The next stage of the dissertation then models that consumption
in MATLAB and finally proposes ways in which to reduce that consumption without
capital investment.
The auxiliary power consumption within Tarong Power Station is recorded by two energy
meters per unit on the main high voltage unit transformer feeds. The energy consumed is
then reported each week as a percentage of unit generation. It is at this high level that the
consumption is currently understood. This project has created energy tracking logic in the
unit control system, a Siemens T3000 installation, to provide additional usage knowledge.
A number of MATLAB models have been produced. The first of these reproduces the
energy usage map of a running unit. The final Simulink model allows modification of the
major component loading to trial energy reduction options. Using this model a reduction
of 10 % at low loads has been achieved. The accuracy of the energy tracking logic and
models created is proven to be within 2 % of the field energy metering.
This dissertation has concluded that meaningful energy efficiency improvement can be
obtained through operational improvement at Tarong Power Station.
University of Southern Queensland
Faculty of Health, Engineering & Sciences
ENG4111/2 Research Project
Limitations of Use
The Council of the University of Southern Queensland, its Faculty of Health, Engineering
& Sciences, and the staff of the University of Southern Queensland, do not accept any
responsibility for the truth, accuracy or completeness of material contained within or
associated with this dissertation.
Persons using all or any part of this material do so at their own risk, and not at the risk of
the Council of the University of Southern Queensland, its Faculty of Health, Engineering
& Sciences or the staff of the University of Southern Queensland.
This dissertation reports an educational exercise and has no purpose or validity beyond
this exercise. The sole purpose of the course pair entitled “Research Project” is to con-
tribute to the overall education within the student’s chosen degree program. This doc-
ument, the associated hardware, software, drawings, and other material set out in the
associated appendices should not be used for any other purpose: if they are so used, it is
entirely at the risk of the user.
Dean
Faculty of Health, Engineering & Sciences
Certification of Dissertation
I certify that the ideas, designs and experimental work, results, analyses and conclusions
set out in this dissertation are entirely my own effort, except where otherwise indicated
and acknowledged.
I further certify that the work is original and has not been previously submitted for
assessment in any other course or institution, except where specifically stated.
Jason Lang
0061014535
Acknowledgments
I would like to make the following acknowledgements;
Stanwell Corporation and in particular David Janes, Jan Potgieter, Les Airs, and Malcolm
Westaway.
USQ and in particular Catherine Hills.
My wife and daughter; Steph and Caitlin Lang.
Without their support and assistance this project would not have became a reality.
Jason Lang
Contents
Abstract i
Acknowledgments iv
List of Figures xi
List of Tables xv
Acronyms & Abbreviations xviii
Chapter 1 Introduction 1
1.1 Project Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.1 Aims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.3 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Overview of the Dissertation . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Chapter 2 Background - Tarong Power Station 6
CONTENTS vii
2.1 History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2 Production Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 Electricity Generation Process . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.4 Unit Energy Requirements - Auxiliary Consumption . . . . . . . . . . . . 14
2.5 Energy Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.6 Control System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Chapter 3 Literature Review 21
3.1 Auxiliary Power Usage and Reduction . . . . . . . . . . . . . . . . . . . . 21
3.2 Power Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.2.1 Average or Real Power . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.2.2 Power Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.2.3 Apparent Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2.4 Reactive Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2.5 Power Triangle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2.6 Wattmeters/Watthour Meters . . . . . . . . . . . . . . . . . . . . . 37
3.3 Practical Industrial Power Measurement . . . . . . . . . . . . . . . . . . . 37
3.4 Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.5 Control Optimisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.6 Power Station Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
CONTENTS viii
Chapter 4 Methodology 46
4.1 Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.2 Initial Data Gathering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.2.1 Drawings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.2.2 Operation and Maintenance Manuals . . . . . . . . . . . . . . . . . 52
4.2.3 Energy Metering Data . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.2.4 T3000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.2.5 Plant Performance Reports . . . . . . . . . . . . . . . . . . . . . . 54
4.3 T3000 Energy Tracking Logic Creation . . . . . . . . . . . . . . . . . . . . 55
4.4 MATLAB Models Creation . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.4.1 Conceptional MATLAB Script Creation . . . . . . . . . . . . . . . 61
4.4.2 SIMULINK Auxiliary Power Model Creation . . . . . . . . . . . . 63
4.4.3 SIMULINK Auxiliary Position Model Creation . . . . . . . . . . . 70
4.4.4 SIMULINK Auxiliary Process Model Creation . . . . . . . . . . . 71
4.4.5 Model Comparison Data . . . . . . . . . . . . . . . . . . . . . . . . 74
4.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Chapter 5 Results and Data Analysis 77
5.1 Performance Reports Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.2 T3000 Daily Auxiliary Total Data . . . . . . . . . . . . . . . . . . . . . . 79
5.3 MATLAB Model Output Data . . . . . . . . . . . . . . . . . . . . . . . . 83
5.3.1 SIMULINK Auxiliary Power Model . . . . . . . . . . . . . . . . . 83
CONTENTS ix
5.3.2 SIMULINK Auxiliary Position Model . . . . . . . . . . . . . . . . 86
5.3.3 SIMULINK Auxiliary Process Model . . . . . . . . . . . . . . . . . 89
5.4 Overall Auxiliary Energy Usage . . . . . . . . . . . . . . . . . . . . . . . . 92
5.5 Other Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5.6 Data Shortcomings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Chapter 6 Proposed Options of Energy Reduction 99
6.1 Induced Draft Fans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
6.2 Forced Draft Fans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
6.3 Primary Air Fans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
6.4 Circulating Water Pumps . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
6.5 Condensate Extraction Pumps . . . . . . . . . . . . . . . . . . . . . . . . 111
6.6 Boiler Feed Pumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
6.7 Reduction Options Summary . . . . . . . . . . . . . . . . . . . . . . . . . 115
6.8 Entire Unit - Optimum Solution . . . . . . . . . . . . . . . . . . . . . . . 117
6.8.1 Low Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
6.8.2 Medium Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
6.8.3 High Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
6.8.4 Reduction Summary . . . . . . . . . . . . . . . . . . . . . . . . . . 121
6.9 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
CONTENTS x
Chapter 7 Conclusion and Further Work 125
7.1 Project Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
7.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
7.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Bibliography 131
Appendix A Project Specification 134
Appendix B Conceptual MATLAB Auxiliary Calculation Script 136
Appendix C Unit Auxiliary Power Consumption Simulink Model - Power
Model 143
Appendix D Unit Auxiliary Power Consumption Simulink Model - Posi-
tion Model 151
Appendix E Unit Auxiliary Power Consumption Simulink Model - Process
Model 159
Appendix F Process Information Management System (PIMS) Extracted
Data 167
Appendix G Located Drawing Data 169
Appendix H Energy Metering Manual Data 170
Appendix I Stanwell Corporation Energy Reductions Recommendations
Report 174
List of Figures
2.1 Monthly Report is a Snapshot of the Data Produced to Track Unit Perfor-
mance (Sands & Blake 2015). . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 Unit 4 T3000 Overview - Screen Capture from Tarong Power Station DCS
(Siemens 2015). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.3 Unit 1 Distribution Overview (Hitachi 2003). . . . . . . . . . . . . . . . . 17
2.4 Station Ancillary Distribution Overview (Hitachi 2003). . . . . . . . . . . 18
3.1 Typical Auxiliary Loads in Coal Fired Power Plant (EPRI 2011). . . . . . 23
3.2 Typical Auxiliary Loads and their Breakdown of Usage in Coal Fired Power
Plant (EPRI 2011). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.3 210MW Coal Fired Power Plant Cycle (Mandi, Rajasheker, Udaykumar &
Yaragatti 2012). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.4 Auxiliary Power Consumption Against Load Factor in Coal Fired Power
Plant (Mandi et al. 2012). . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.5 The Power Triangle relates Apparent, Reactive and Average Power with
Power Factor (Boylestad 2003). . . . . . . . . . . . . . . . . . . . . . . . . 36
3.6 The Simplified Control Scheme of a Coal-fired Power Plant (Simon, Stoica,
Rodriguez-Ayerbe, Dumur & Wertz 2010). . . . . . . . . . . . . . . . . . . 42
LIST OF FIGURES xii
4.1 Initial Test Real Power Calculation Logic - Screen Capture from Tarong
Power Station DCS (Siemens 2015). . . . . . . . . . . . . . . . . . . . . . 57
4.2 Unit 4 A Unit Transformer Energy Tracking Logic - Screen Capture from
Tarong Power Station DCS (Siemens 2015). . . . . . . . . . . . . . . . . . 58
4.3 Unit 4 A and C Energy Tracking Logic - Screen Capture from Tarong
Power Station DCS (Siemens 2015). . . . . . . . . . . . . . . . . . . . . . 60
4.4 Initial SIMULINK model of Induced Draft Fan - Basic Concept Version. . 64
4.5 Initial SIMULINK model of Induced Draft Fan - Lookup Table Concept
Version. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.6 Auxiliary SIMULINK Model - Power Version of Full Unit - A Bus Common
Calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4.7 Auxiliary SIMULINK Model - Power Version of Full Unit - A Bus Load
Calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.8 Auxiliary SIMULINK Model - Position Version of Full Unit - A Bus Load
Calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.9 Auxiliary SIMULINK Model - Process Version of Full Unit - A Bus Load
Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.1 A Comparison of the Typical Auxiliary Load Distribution in Coal Fired
Power Plant to the Average Tarong Auxiliary Load (EPRI 2011). . . . . . 81
5.2 Tarong Power Station Baseline Auxiliary Energy Consumption - Power
Model Plant Totals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.3 Tarong Power Station Baseline Auxiliary Energy Consumption - Position
Model Plant Totals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.4 Tarong Power Station Baseline Auxiliary Energy Consumption - Process
Model Plant Totals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
LIST OF FIGURES xiii
5.5 Tarong Power Station Baseline Auxiliary Energy Consumption - Process
Model versus Formula. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.6 A Boiler Feed Pump Curve with the Pump Loading Shown in Realtime -
Unit 4A BFP 350MW (Siemens 2015). . . . . . . . . . . . . . . . . . . . . 95
5.7 Load Profiles of Unit 4 Tarong Power Station. Top Left, 365MW-190MW.
Top Right, 365MW-220MW. Bottom Left, 140MW. Bottom Right, 140-
200MW (Siemens 2015). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
6.1 Air and Gas System Components - Tarong Power Station (Siemens 2015). 101
6.2 Circulating Water System Components - Tarong Power Station (Siemens
2015). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.3 Boiler Feed Pump System Components - Tarong Power Station (Siemens
2015). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
6.4 Boiler Feed Pump Plant Area - Process Model versus Reduction Options . 116
6.5 Full Unit Auxiliary Energy Consumption Comparison No 1 - Baseline ver-
sus Risk Reduction Options. . . . . . . . . . . . . . . . . . . . . . . . . . . 122
6.6 Full Unit Auxiliary Energy Consumption Comparison No 2 - Baseline ver-
sus Risk Reduction Options. . . . . . . . . . . . . . . . . . . . . . . . . . . 123
C.1 Unit Auxiliary Power Consumption Simulink Model - Power Model Part 1 144
C.2 Unit Auxiliary Power Consumption Simulink Model - Power Model Part 2 145
C.3 Unit Auxiliary Power Consumption Simulink Model - Power Model Part 3 146
C.4 Unit Auxiliary Power Consumption Simulink Model - Power Model Part 4 147
C.5 Unit Auxiliary Power Consumption Simulink Model - Power Model Part 5 148
C.6 Unit Auxiliary Power Consumption Simulink Model - Power Model Part 6 149
C.7 Unit Auxiliary Power Consumption Simulink Model - Power Model Part 7 150
LIST OF FIGURES xiv
D.1 Unit Auxiliary Power Consumption Simulink Model - Position Model Part 1152
D.2 Unit Auxiliary Power Consumption Simulink Model - Position Model Part 2153
D.3 Unit Auxiliary Power Consumption Simulink Model - Position Model Part 3154
D.4 Unit Auxiliary Power Consumption Simulink Model - Position Model Part 4155
D.5 Unit Auxiliary Power Consumption Simulink Model - Position Model Part 5156
D.6 Unit Auxiliary Power Consumption Simulink Model - Position Model Part 6157
D.7 Unit Auxiliary Power Consumption Simulink Model - Position Model Part 7158
E.1 Unit Auxiliary Power Consumption Simulink Model - Process Model Part 1 160
E.2 Unit Auxiliary Power Consumption Simulink Model - Process Model Part 2 161
E.3 Unit Auxiliary Power Consumption Simulink Model - Process Model Part 3 162
E.4 Unit Auxiliary Power Consumption Simulink Model - Process Model Part 4 163
E.5 Unit Auxiliary Power Consumption Simulink Model - Process Model Part 5 164
E.6 Unit Auxiliary Power Consumption Simulink Model - Process Model Part 6 165
E.7 Unit Auxiliary Power Consumption Simulink Model - Process Model Part 7 166
F.1 Example of PIMS Data Extracted for the 7th of July 2015 (Siemens 2015) 168
H.1 List of Energy Meters Installed at Tarong Power Station (Godsmark 2001) 171
H.2 Drawing of Typical Energy Meter Installation at Tarong Power Station
(Godsmark 2001) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
H.3 Example of Typical Energy Meter PLC Registers and Functions (Godsmark
2001) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
I.1 Part 1 of Energy Reduction Recommendations Memorandum to David Janes175
LIST OF FIGURES xv
I.2 Part 2 of Energy Reduction Recommendations Memorandum to David Janes176
List of Tables
2.1 Cost Contribution of Major Factors to the Cost of Producing 1 MW (Breda
2014). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Major Drives as summarised from Hitachi Operations and Maintenance
Manuals (Hitachi 2003). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.1 Auxiliary Energy Usage Figures in Percentage by Unit for the Last 10
Months (Sands & Blake 2015). . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2 A Sample of Auxiliary Energy Consumption figures from Around the World.
(EPRI 2011) (Mandi et al. 2012) (Boveri 2015) (Sands & Blake 2015) . . 24
4.1 Auxiliary Load - Detail of all Loads that Contribute to the Auxiliary Power
Consumption of the Unit (Alstrom 1981). Note N/A (Not Available) Power
Factors were Unable to be Located. . . . . . . . . . . . . . . . . . . . . . . 51
4.2 Auxiliary Energy Usage Figures Comparison between Plant Energy Meter,
Tracking Logic and the MATLAB Script Output. Totals are in MW/hr
and the Percentage Error in Presented in Brackets. . . . . . . . . . . . . . 61
4.3 MATLAB Script Output - Auxiliary Energy Usage Figures at the 13 Load
Points. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.4 Auxiliary Energy Usage Figures Comparison between the MATLAB Script
Output and the Power Based Model Output at 190 MW. Totals are in
MWhrs and the Percentage Error in Presented in Brackets. . . . . . . . . 67
LIST OF TABLES xvii
4.5 Assessment of Auxiliary Load Drive Elements, Plant Configuration and
there Capacity for Load Share. . . . . . . . . . . . . . . . . . . . . . . . . 70
4.6 Auxiliary Loads and their Identified Process Measurement. . . . . . . . . 73
4.7 Auxiliary Energy Usage Figures Comparison between the Three SIMULINK
Models at 190MW. Totals are in MWhrs. . . . . . . . . . . . . . . . . . . 74
5.1 Auxiliary Energy Usage Figures - Unit 4 for the Last 10 Months (Sands &
Blake 2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.2 Auxiliary Energy Usage Figures in MWhrs - Pulverisers - Output from
T3000 Energy Tracking Logic Daily Totals. . . . . . . . . . . . . . . . . . 80
5.3 Unit 4 Auxiliary Energy Usage Figures in MWhrs - All Plant Areas -
Output from T3000 Energy Tracking Logic Daily Totals. . . . . . . . . . . 82
5.4 Auxiliary Energy Usage Figures in MWhrs - Plant Area Totals from SIMULINK
Power Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
5.5 Auxiliary Energy Usage Figures per Drive in MWhrs - Output from SIMULINK
Power Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.6 Auxiliary Energy Usage Figures in MWhrs - Plant Area Totals from SIMULINK
Position Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.7 Auxiliary Energy Usage Figures per Drive in MWhrs - Output from SIMULINK
Position Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
5.8 Auxiliary Energy Usage Figures in MWhrs - Plant Area Totals from SIMULINK
Process Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.9 Auxiliary Energy Usage Figures per Drive in MWhrs - Output from SIMULINK
Process Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
6.1 MATLAB Process Model Output - Induced Draft Fan System Energy Us-
age Figures at 9 Load Points to be used as a Baseline. Generator Load Set
Point in MW and Energy Totals in MWhrs. . . . . . . . . . . . . . . . . . 100
LIST OF TABLES xviii
6.2 MATLAB Process Model Output - Induced Draft Fan System Energy Us-
age Figures at 9 Load Points after Reduction Option 1 Implemented. Gen-
erator Load Set Point in MW and Energy Totals in MWhrs. . . . . . . . . 102
6.3 MATLAB Process Model Output - Induced Draft Fan System Energy Us-
age Figures at 9 Load Points after Reduction Option 2 Implemented. Gen-
erator Load Set Point in MW and Energy Totals in MWhrs. . . . . . . . . 103
6.4 MATLAB Process Model Output - Induced Draft Fan System Energy Us-
age Figures at 9 Load Points after Reduction Option 3 Implemented. Gen-
erator Load Set Point in MW and Energy Totals in MWhrs. . . . . . . . . 103
6.5 MATLAB Process Model Output - Forced Draft Fan System Energy Usage
Figures at 9 Load Points to be used as a Baseline. Generator Load Set Point
in MW and Energy Totals in MWhrs. . . . . . . . . . . . . . . . . . . . . 104
6.6 MATLAB Process Model Output - Forced Draft Fan System Energy Usage
Figures at 9 Load Points after Reduction Option 1. Generator Load Set
Point in MW and Energy Totals in MWhrs. . . . . . . . . . . . . . . . . . 105
6.7 MATLAB Process Model Output - Forced Draft Fan System Energy Usage
Figures at 9 Load Points after Reduction Option 2. Generator Load Set
Point in MW and Energy Totals in MWhrs. . . . . . . . . . . . . . . . . . 106
6.8 MATLAB Process Model Output - Primary Air Fan System Energy Usage
Figures at 9 Load Points to be used as a Baseline. Generator Load Set
Point in MW and Energy Totals in MWhrs. . . . . . . . . . . . . . . . . . 106
6.9 MATLAB Process Model Output - Primary Air Fan System Energy Usage
Figures at 9 Load Points after Implementation of Option 1. Generator
Load Set Point in MW and Energy Totals in MWhrs. . . . . . . . . . . . 107
6.10 MATLAB Process Model Output - Primary Air Fan System Energy Usage
Figures at 9 Load Points after Implementation of Option 2. Generator
Load Set Point in MW and Energy Totals in MWhrs. . . . . . . . . . . . 108
LIST OF TABLES xix
6.11 MATLAB Process Model Output - Circulating Water System Energy Usage
Figures at 9 Load Points to be used as a Baseline. Generator Load Set Point
in MW and Energy Totals in MWhrs. . . . . . . . . . . . . . . . . . . . . 110
6.12 MATLAB Process Model Output - Circulating Water System Energy Usage
Figures at 9 Load Points after Option 1 Implemented. Generator Load Set
Point in MW and Energy Totals in MWhrs . . . . . . . . . . . . . . . . . 110
6.13 MATLAB Process Model Output - Circulating Water System Energy Usage
Figures at 9 Load Points after Option 2 Implemented. Generator Load Set
Point in MW and Energy Totals in MW/hr . . . . . . . . . . . . . . . . . 111
6.14 MATLAB Process Model Output - Condensate Extraction System Energy
Usage Figures at 9 Load Points to be used as a Baseline. Generator Load
Set Point in MW and Energy Totals in MWhrs. . . . . . . . . . . . . . . . 112
6.15 MATLAB Process Model Output - Boiler Feed Pump System Energy Usage
Figures at 9 Load Points to be used as a Baseline. Generator Load Set Point
in MW and Energy Totals in MWhrs. . . . . . . . . . . . . . . . . . . . . 112
6.16 MATLAB Process Model Output - Boiler Feed Pump System Energy Usage
Figures at 9 Load Points after Option 1 Implemented. Generator Load Set
Point in MW and Energy Totals in MWhrs. . . . . . . . . . . . . . . . . . 114
6.17 MATLAB Process Model Output - Boiler Feed Pump System Energy Usage
Figures at 9 Load Points after Option 2 Implemented. Generator Load Set
Point in MW and Energy Totals in MWhrs. . . . . . . . . . . . . . . . . . 115
6.18 Impact of Reduction Options in Each Plant Area at Different Generator
Load Set Points in MW - Resulting Impact Presented as a Percentage. . . 117
6.19 Impact of Reduction Options on the Overall Auxiliary Energy Consump-
tion at Different Generator Load Set Points in MW - Resulting Impact
Presented as a Percentage. . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
6.20 MATLAB Process Model Output - Full Unit Auxiliary Energy Usage Base-
line Figures at 9 Load Points. Generator Load Set Point in MW and Energy
Totals in MWhrs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
LIST OF TABLES xx
6.21 MATLAB Process Model Output - Full Unit Energy Usage Figures at 9
Load Points after Low Risk Options are Implemented. Generator Load Set
Point in MW and Energy Totals in MWhrs. . . . . . . . . . . . . . . . . . 119
6.22 MATLAB Process Model Output - Full Unit Energy Usage Figures at 9
Load Points after Low and Medium Risk Options are Implemented. Gen-
erator Load Set Point in MW and Energy Totals in MWhrs. . . . . . . . . 120
6.23 MATLAB Process Model Output - Full Unit Energy Usage Figures at 9
Load Points after Low, Medium and High Risk Options are Implemented.
Generator Load Set Point in MW and Energy Totals in MWhrs. . . . . . 121
G.1 A Sample of Drawings Located during the Data Gathering Stage (Hitachi
2003). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Acronyms & Abbreviations
AEMO Australian Electricity Market Operator
AP Automation Processor
ASD Adjustable Speed Drives
AVR Automatic Voltage Regulation
BFP Boiler Feed Pump
CEP Condensate Extraction Pump
CPU Central Processing Unit
CT Current Transformer
CW Circulating Water
DA Deaerator
DC Direct Current
DCS Distributed Control System
DF Displacement Power Factor
DOL Direct on Line
EPRI Electric Power Research Institute
ESP Electrostatic Precipitator
FD Forced Draft
FIFO First In First Out
GOC Government Owned Corporation
GT Gas Turbine
HMI Human Machine Interface
HP High Pressure
HP5 HP heater 5
Acronyms & Abbreviations xxii
HP6 HP heater 6
hr Hour
HV High Voltage
I&C Instrumentation and Control
ID Induced Draft
IEEE Institute of Electrical and Electronics Engineers
I/O Inputs/Outputs
MV Medium Voltage
IP Intermediate Pressure
JSEA Job Safety and Environmental Analysis
kW Kilowatt
LP Low Pressure
MV Medium Voltage
MW Megawatt
MWhr Megawatt hour
NGERS National Greenhouse Energy Reporting Scheme
NOX Nitrite Oxide
PID Proportional Integral Derivative
P&ID Process and Instrumentation Drawing
PA Primary Air
PF Pulverised Fuel
PF Power Factor
PIMS Process Information Management Solutions
PLC Programmable Logic Controllers
QEC Queensland Electricity Commission
SCADA Supervisory Control and Data Acquisition
SISO Single Input Single Output
Stanwell Stanwell Corporation
Tarong Tarong Power Station
THD Total Harmonic Distortion
VT Voltage Transformer
Chapter 1
Introduction
This chapter aims to introduce the project, its drivers, its key aims, objectives and how
they are proposed to be achieved.
In an ever more energy hungry austere world the way in which energy is produced and
consumed is under constant scrutiny. Every Megawatt (MW) produced in a coal fired
power station consumes valuable finite resources, coal and oil. The production of electric-
ity emits pollution into the atmosphere leaving an indelible footprint on the landscape
and atmosphere. This carbon footprint, as it is known, that the Tarong Power Station
(TPS) leaves is larger than most energy consumers in Australia. TPS is ranked in the
top ten Australian polluters and as such is always under pressure to improve its envi-
ronmental performance. TPS is run by Stanwell Corporation (Stanwell), a Government
Owned Corporation (GOC). As such it is owned by the people of Queensland and while
environmental performance is important so is financial stainability. Couple these two
factors with global downturns, market oversupply and low price in the electricity market
you have a business that is always actively looking to reduce costs. Time and energy
is spent to improve fuel burn returns, reduce water usage, implement reduced manning
practices, improve market share and reduce the cost of maintenance work. One of the
last major cost influences, auxiliary power consumption, has only been lightly touched on
during the recent stakeholder driven cost reduction exercises. It is this cost of internal
energy consumed through auxiliary power consumption that is the focus of this project.
1.1 Project Outline 2
1.1 Project Outline
This section outlines some further detail on the issue of auxiliary power consumption
to give context to the project direction. It will also detail the projects aims, scope and
objectives.
Auxiliary energy consumption is a large contributor to the cost of power production.
Within Tarong Power Station it is not fully understood, that is, it is understood to a basic
level. It is known that of the energy produced at TPS approximately 10 % is required to
run in house auxiliaries and therefore cannot be exported for sale. This auxiliary energy
usage is a significant cost to the business and adds to the cost of producing one MW of
electricity. This auxiliary power usage is metered on each unit and is reported on each
month. If a unit has an increased energy usage at the end of the month not only is it
accepted and unchallenged but there is no easy way to understand the underlying cause.
Therefore there is no way to prevent a future high energy usage occurrence of the same
nature. Since the usage is metered at a high level, unit level, there is also no reasonable
way to assess the usage and implement changes to the plant to reduce the usage. It can
also be noted that the energy usage whilst visible is generally lost in the other concerns
of the station.
The major contributors to the auxiliary energy consumption are already broadly known
and loosely understood. It is the plants large high voltage (HV) motors which consume
a huge amount of electricity to produce the motive forces required by the electricity
generation process. Reducing this consumption via major plant upgrade would require
enormous capital investment and plant redesign. Further to this it is widely accepted that
the current level of unit energy consumption, as part of the designed efficiency figure, is
the minimum achievable usage. It is essentially viewed as unavoidable byproduct of the
process requirements. The plant was designed in the mid 1970’s with a strong reliability
focus. To ensure high plant availability the plant design included many redundant and
oversized drives. The station was also designed to be a base load 350 MW unit. The
energy consumption across the full load range of the plant was not the primary focus of
the design phase of the station. Peak efficiency at 350 MW was the primary aim of the
station design.
1.1 Project Outline 3
1.1.1 Aims
The questions that this project seeks to address are two stage, firstly can the auxiliary
energy consumption of unit at Tarong Power Station be fully understood? Secondly can
this energy usage be reduced solely through control optimisation? That is to reduce the
auxiliary energy consumption without capital investment and solely by control system
logic modification. This leads to the creation of the aims of this project. The first
primary project aim is to better understand the auxiliary energy consumption at Tarong
Power Station. Once understood the second primary aim of reducing the usage through
control optimisation can be addressed.
In building the tools to understand the stations auxiliary usage the project will create
valuable data in T3000. The created logic will be able to monitor and report on usage
for the project and into the future. This had led to a secondary aim of leaving in place
lasting a energy information source.
These aims are summarised here for reference.
Primary Aim 1 To understand the auxiliary consumption within Tarong Power Sta-
tions 350 MW coal fired units;
Primary Aim 2 Identify potential control based energy reduction strategies;
Secondary Aim 1 To leave in place energy consumption information that the station
can utilise into the future.
Notes The primary aims are those that drive the project objectives. The secondary
aims are those that will be achieved through the primary aims.
1.1.2 Objectives
The project objectives drive how the project aims will be achieved. This leads to the
project and the associated dissertation having four primary objectives. The first is to
understand the associated project topics more deeply and uncover any existing work that
can be used to assist the completion of this project. This will be achieved through a wide
ranging literature review. The next two objectives will be to model the auxiliary energy
usage of a 350 MW coal fired unit using MATLAB and to create energy tracking logic in
1.1 Project Outline 4
the T3000 Distributed Control System (DCS). Using this model and T3000 data the final
primary objective is to propose control changes to reduce the internal electrical energy
consumed to produce electricity. Proposed changes will be implemented in the MATLAB
model to confirm the energy impact. The secondary objectives of this project are to
utilise the T3000 simulator and the TPS generating plant to implement the proposed
control changes. The achievement of these secondary objectives will depend heavily on
the outcomes of the first stages and the risk assessment of any proposed changes. These
secondary aims are an extension to the primary stated objectives and will be proven
theoretically if implementation is not possible.
Primary Objective 1 Complete a literature review of issues pertinent to the project;
Primary Objective 2 Create a SIMULINK model, in MATLAB, of the energy profile
of a single unit;
Primary Objective 3 Create energy tracking logic in the T3000 Digital Control Sys-
tem;
Primary Objective 4 Identify potential control based energy reduction strategies and
propose changes to the operational plant;
Secondary Objective 1 Trial energy reduction options on the T3000 Unit simulator;
Secondary Objective 2 Implement energy reduction strategy on an operational Unit.
Note Primary objectives are those that drive this project and the secondary objectives
are extension tasks if time and plant risk assessment permit.
1.1.3 Scope
The scope of this project is limited to the Unit Auxiliary energy usage. Specifically
the energy usage on the 350 MW coal fired units location at Tarong Power Station,
Queensland, Australia. The Power Station Ancillary energy usage ie Coal Handling Plant,
Tarong North, Administration areas and Meandu mine energy usage is excluded from this
project.
1.2 Overview of the Dissertation 5
1.2 Overview of the Dissertation
This dissertation is organised as follows:
Preface has a list of abbreviations that will be used throughout the dissertation;
Chapter 1 is an introduction to the problem of auxiliary energy usage in coal fired power
stations;
Chapter 2 is a background information on the various facets of this project;
Chapter 3 is a discussion of the existing knowledge on the dissertation topic and is
presented through discussion of existing literature;
Chapter 4 discusses the methods utilised in the completion of this project;
Chapter 5 presents the results of the created T3000 logic and MATLAB models;
Chapter 6 analyses and evaluates the data collected from all sources;
Chapter 7 discusses the reduction options and presents the results of these options im-
plemented using the MATLAB models;
Chapter 8 concludes the dissertation with a concise presentation of the project achieve-
ments and suggests further work that could be undertaken in the area of ‘energy
modeling and reduction in the power station environment’.
1.3 Conclusion
In this chapter the project overview was explored from many perspectives to allow a
solid understanding of the drivers for this project, its aims and the final outcomes it will
produce. The next chapter will provide the background information required on auxiliary
power usage within Tarong Power Station to allow the reader the basis on which to better
understand the future project outputs.
Chapter 2
Background - Tarong Power
Station
This chapter is to provide background information that relates to the topic. The majority
of the information within this chapter is a summary of the Operations and Maintenance
manuals for the Tarong Power Station plant (Hitachi 2003). Further detail is based on the
experience of the project author Jason Lang who has 20 years experience in the electricity
generation industry (Lang 2015).
2.1 History
This section outlines the brief history of Tarong Power Station from construction through
to current day.
Tarong Power Station construction started in 1979 after it was commissioned to be built
as part of 10 coal fired units in Queensland by then Queensland Premier Sir Joh Bjelke-
Peterson. Tarong had four 350 MW units fully commissioned by November 1986 with Unit
1 in full operation during 1981. These units were designed to be base load for Queensland
for 25 years. The units design efficiency was 28 % at 350 MW and were installed with
the best technology available at the time (Brady 1996) (Thiess 2015).
Move forward to the modern day, Tarong is past its design life and there have been
many evolutions in the Queensland Electricity Industry. Larger, more efficient units have
2.2 Production Costs 7
been built, the industry is partly deregulated and gas is now a significant factor in the
marketplace. Tarong has established itself in the middle of the Australian Electricity
Market Operator (AEMO) trading dispatch order of merit. The Queensland electricity
market has been depressed with state load reducing, an oversupply and solar all impacting
price and load profiles. To extend the station life and to remain a competitive force Tarong
Power Station has undergone a major capital works program. These include a full control
systems refit, field device upgrade, burner front upgrade to low Nitrite Oxide (NOX),
major turbine refurbishment program and generator rewinds (Australia 2015).
The half-life refit work has significantly improved the underlying long term life on the plant
but has only slightly improved the efficiency factor. TPS is also no longer a base load
station with the plant being regularly requested to load ramp. This mode of operation is
outside of the original plant design specification and therefore not the most efficient plant
operation. To return to base load the overall cost to produce one MW of energy needs to
be reduced to allow the plant to be bid in lower in the market dispatch order.
2.2 Production Costs
This section outlines the cost of producing one MW of electricity at TPS. The cost of
this electricity production process has many contributing factors. Tarong Power Station’s
major contributing factors are the cost of fuel, water, overhaul maintenance, auxiliary
energy consumption, ancillary energy consumption and fixed overheads. The estimate of
the breakdown of the cost contributions to each MW are shown in table 2.1.
Cost Contributor Contribution
Fuel 30 %
Water 30 %
Overhaul Maintenance 20 %
Auxiliary Energy Usage 15 %
Ancillary Energy Usage 4 %
Fixed Overheads 1 %
Table 2.1: Cost Contribution of Major Factors to the Cost of Producing 1 MW (Breda 2014).
The fuel and water costs are fixed by long term contract and therefore cannot be readily
2.2 Production Costs 8
reduced. However the amount of fuel and water consumed can be reduced through effi-
cient practices. Overhaul costs are strongly linked to long term asset strategy which are
locked down by multiple year strategy. Moves to reduce this overhaul cost are possible
but hold a high level of risk to plant performance and reliability. Auxiliary energy usage
costs are those directly associated with the energy required to generate power. This power
usage is wide spread from control system power requirements, through to High Voltage
Drives and generator excitation. This cost is considered to be fixed during the initial
design process but is monitored. Ancillary energy usage costs are those associated with
common processes like coal handling plant, air compression, demineralised water produc-
tion, chemical plants and water distribution. The fixed costs are those support items
that a company generating electricity requires. All manner of costs make up this factor
including head office, payroll, market and trading, procurement, operations, maintenance,
and engineering. These fixed costs are a small percentage of the overall cost of 1 MW but
are relatively easily influenced by plant management. They are therefore under constant
scrutiny and downward pressure (Economy 2013).
Key parameters are reported on each month to track the items that affect the amount
and cost of the saleable product, in this case electricity exported to the grid. These key
parameters align with the key cost contributors above. Figure 2.1 is a sample of the data
presented in monthly reports.
The data shows, per unit, the energy and performance results for the month. While they
are all impacted by the unit load profile they also indicate other unit qualities which are
explored briefly below. Megawatt hours (MWhr) produced or generated energy exported
is metered at the 275 kV level as the power is leaving the unit transformer to meet the
grid. Auxiliary energy usage is a measure of the electricity used to produce the MWhrs
exported in the month. It is metered at the unit 20 kV to 6.6 kV transformers that feed
the majority of energy consuming devices on the units. Capacity factor is an indication of
how the plant is loaded in relation to its full capacity. Efficiency and heat rate shows how
effectively the coal consumed is utilised. It is a combination of many factors including coal
quality, ash content, moisture content, mill grind quality, air leakage etc. Availability is
the measure, in percentage, of the amount of time the unit is available to meet the
demands of the market. Availability is impacted by breakdowns, plant trips and outages.
Solid long term maintenance practices and underlying plant design influence this figure.
The grid and export requirements are managed by Australian Electricity Market Opera-
2.2 Production Costs 9
Figure 2.1: Monthly Report is a Snapshot of the Data Produced to Track Unit Performance
(Sands & Blake 2015).
2.3 Electricity Generation Process 10
tors (AEMO). The unit load, its available capacity and metered energy is communicated
in real time to AEMO. The station transformer’s usage is measured and read by Ergon
Energy and Stanwell is billed for its use. The station transformer energy is used to run the
ancillary services of the station like administration buildings, coal handling plant drives,
air compression and water treatment. The station boards also link to each unit to assist
in unit start-up until it can sustain its own load. Tarong also has a 15 MW Gas Turbine
(GT) connected to the station switchboards which is used for start-up power during black
start operations. This is a situation when the state grid is unable to supply the station
power. The GT can also be used to generate back to the grid. All metering is audited
and certified by a third party at routine intervals to ensure transparency.
2.3 Electricity Generation Process
This section outlines the power generation process or what is known as the cycle. The
Tarong Power Station is a closed cycle regeneration sub critical thermal unit utilising coal
as its primary fuel source. The cycle consists primarily of 11 major components. A broad
understanding of these components will assist to better understand the future discussion
this project will undertake.
Cooling Water Cycle: Used to convert the steam back to condensate once it has
passed through the low pressure turbine, heat absorbed is removed by the associated
cooling towers.
Condensate and Feedwater System: Collects the condensate and transports it through
the heater stages to the boiler drum;
Steam Cycle: This cycle converts the preheated feedwater to steam and superheats it
to precise pressure and temperature require by the turbine. The steam cycle also
includes supplying heat to the low and high pressure heaters.
Turbine System: Produces the mechanical motive force required to drive the generator
through conversion of steam energy;
Generator System: Produces power via a 6 pole generator with a stationary stator
and rotating DC field. The generator output is 20 kV which is stepped up for
transmission and stepped down for internal unit auxiliary power usage.
2.3 Electricity Generation Process 11
Fuel System: This system supplies the fuel for the boiler flame via oil supply, coal
pulverisers and associated burner fronts.
Boiler: Produces the heat required for steam generation and consists of multiple stages
of tubing.
Air and Gas: Forced Draft, Induced Draft, Primary Air and Air Heaters all combine
with dampers to supply the necessary air for the mills and boiler operation.
Control System: Servers, processors, control cabinets, field power supplies, field de-
vices, unit control desks and operator workstations that are required to monitor
and control the unit.
HV/LV Distribution; Switchboards, cabling and circuit breakers that control the dis-
tribution of electricity for the unit;
Chemical: Chemical instrumentation, wetracks, chemical dosing, and condensate pol-
ishing plants required to monitor and maintain the unit cycle chemistry.
Figure 2.2 is a realtime capture of the TPS Unit 4 HMI overview page which shows the
major processes that combine to generate power in a thermal unit.
The cycle is considered closed as once the steam energy is utilised by the Low Pressure
(LP) turbine it is then recondensed and used in the cycle again. Fresh demineralised
water is only required to be added to the system to make up for losses. The paragraphs
below explain the cycle operation in more detail.
The Cooling Water, Condensate and Feedwater Cycle
Cooling water is used in the condenser to recondense the steam into water. The cooling
water picks up heat during this process; this heat is then removed by circulation through
the natural draft cooling towers. This circulation is achieved by Circulating Water (CW)
Pumps. The condensed steam is then considered condensate and is extracted from the
condenser hotwell by the Condensate Extraction Pumps (CEPs). The CEPs transport
the condensate through the gland seal condenser and three low pressure heaters. The
condensate picks up heat as it passes through each heater. This heat is supplied by
gland seal steam and steam picked off the LP turbine stages. Once the condensate has
2.3 Electricity Generation Process 12
Figure 2.2: Unit 4 T3000 Overview - Screen Capture from Tarong Power Station DCS
(Siemens 2015).
2.3 Electricity Generation Process 13
passed through the LP heaters it arrives in heater 4 or a vessel better known as the
Deaerator (DA). It is a known as the Deaerator as part of its function is to remove any
CO2 from the condensate as it enters the vessel. The Boiler Feed Pumps (BFPs) draw
their feed water from the DA then transport it through two high pressure (HP) heaters
and economiser tubing into the boiler drum. Through this part of the cycle the water is
known as feedwater and picks up further heat. The heat for HP heater 5 (HP5) and 6
(HP6) is supplied by steam tapped off the higher temperature Intermediate Pressure (IP)
turbine stage and the cold reheat line respectively. The economiser tubing heat is supplied
by the waste heat extracted from the boiler air as it makes its way to the chimney stack.
From the boiler drum the feedwater is converted to saturated steam and enters the steam
cycle. The steam builds enthalpy as it works its way through the primary, secondary and
tertiary superheater stages to arrive at the HP turbine at 535◦C and 17 MPa. The steam
drops pressure and temperature as its stored energy is imparted on the HP turbine. The
steam is then fed back into the boiler as cold reheat steam to boost the pressure and
temperature again in the reheat boiler stages. Once the temperature has reached 535◦C
and 3.8 MPa it returns as hot reheat steam to drive through the intermediate and low
pressure turbine into the condenser to complete the cycle.
The Boiler, Air and Gas Cycle
The boiler flame is supported by the fuel, air, and gas systems. The air and gas cycle
consists of two Forced Draft (FD) Fans, two Induced Draft (ID) Fans, two Primary Air
(PA) Fans, two Air Heaters, Electrostatic Precipitator (ESP) and the chimney stack. This
air and gas system supplies air to the boiler and pulverisers. The fuel system consists of
six coal pulverisers and the interconnecting pipework to take the Pulverised Fuel (PF)
to the six burner fronts. There is 3 to 5 of these 6 pulverisers in service at any point
in time depending on load requirements. The Tarong pulverisers are ball grinding mills
that grind the coal to fine dust to be transported by the 100◦C PA to the burner front.
The PF enters the boiler and is lit off by either the existing flame or by gas ignition
supported oil flame. The rate of PF added supports the floating boiler flame to provide
the required heat to the boiler tubing transporting the steam. The boiler requires large
amounts of air to feed the flame and to carry away ash from the combustion process. This
air is supplied from two FD fans drawing in fresh air and two ID fans drawing air out
through the precipitator then expelling it through the chimney stack. The precipitators
2.4 Unit Energy Requirements - Auxiliary Consumption 14
are located between the boiler and the stack. The ESP removes particulate matter via
attraction to plates with pulsed direct current (DC).
Minor Auxiliary Systems
There are many other smaller systems that contribute to the Tarong thermal cycle. Below
are some of the smaller systems;
Condensate Polisher: Chemically cleans the unit cycle water;
Flame Detector Cooling Fans: Supplies cooling air to burner front flame detectors;
Oil Systems: Supplies oil for bearings, hydraulic equipment and burner front ignition
needs;
Coal Feeders: Controls the feed rate of coal into pulverisers;
Stator Cooling Water: Supplies water for generator cooling.
2.4 Unit Energy Requirements - Auxiliary Consumption
This section outlines the major items of energy use within the unit cycle that are deemed
to be auxiliary. The major energy consuming devices are those associated with the major
functions of the unit. They are condensate and feedwater, fuel supply, air supply, cooling
water and generator excitation. Primarily these are large Direct on Line (DOL) 6.6 kV
motor driven devices.
The functions of the units boiler are to combine the fuel, air and heat required to sustain
combustion thus providing heat to the boiler tubing. This combustion process requires
large amounts of air which is supplied by two 2.95 MW 6.6 kV motor driven Forced Draft
Fans through an air heater into the boiler. The fuel is supplied to the boiler by coal
pulverisers or mills. Each mill is driven by a 200 kW 415 V motor to drive the grinding
table via a speed reduction gearbox. The coal is supplied into each mill via a coal feeder
driven by a smaller 4 kW variable speed motor. Once ground the coal is picked up and
transported to the boiler by the air supplied by two 0.9 MW 6.6 kV motors that drive
the Primary Air Fans. Some of this air is heated by the air heater, hot PA, and some
2.4 Unit Energy Requirements - Auxiliary Consumption 15
is cold PA. The hot and cold PA are then combined to ensure the optimal combustion
temperature is maintained. All of this air needs to be removed from the boiler at the
correct rate to ensure a vacuum is always maintained in the boiler enclosure. This is
achieved by two 4.55 MW motors that drive the Induced Draft Fans. These fans draw
air and non-combusted waste material out of the boiler through the air heater and over
the economiser tubing. The air then passes through the electrostatic precipitator, which
remove particulate matter before discharging the air out of the chimney. The boiler tubing
requires water which is converted to driving steam energy of a thermal sub critical power
plant. This boiler water is fed up to the boiler drum by pump units in two stages. During
the first stage water is extracted from the hotwell with one of two 1.1 MW 6.6 kV motor
driven condensate extraction pumps. The CEP then pushes the water through the LP
heaters into the DA. The second stage is two of the three 5.65 MW 6.6kV motor driven
boiler feed pumps deliver the feedwater at the correct rate through the economiser into
the boiler drum. The rate of feedwater is controlled by the variable speed nature of the
pump section of the BFPs. An actuated scoop element controls the oil flow inside the
fluid coupling to link the DOL motor to the pump element. Cooling water is circulated
through the condenser and out to the cooling tower for heat removal. This is achieved by
two 1.35 MW 6.6 kV motor driven circulating water pumps. The final large consumer of
energy is the generator excitation system. This system uses an excitation transformer to
reduce the 6.6 kV to 700 V AC which in turn is fed into the Automatic Voltage Regulation
(AVR) system to convert to high current DC. The current level of this DC is varied and
applied to the rotor of the generator (Hitachi 2003).
Table 2.2 is a listing of the major drives and their ratings as discussed in the section
above.
Many other smaller loads contribute to the total auxiliary power usage of a unit including
control system power supplies, flame detection cooling fans, lube oil pumps, stator cooling
water pumps, air heater drive motors, polisher booster pumps and deaerator booster
pumps.
2.5 Energy Monitoring 16
Drive Rating
Force Draft Fan 2.95 MW
Induced Draft Fan 4.55 MW
Primary Air Fan 0.9 MW
Boiler Feed Pump 5.65 MW
Circulating Water Pump 1.35 MW
Condensate Extraction Pump 1.1 MW
Automatic Voltage Regulator 1.1 MW
Table 2.2: Major Drives as summarised from Hitachi Operations and Maintenance Manuals
(Hitachi 2003).
2.5 Energy Monitoring
This section outlines the way in which energy has been and is currently monitored on the
Tarong Power Station site.
In the year 2000 the station was approaching the end of its design life. It was still part
of the Queensland Electricity Commission (QEC) and a move to a deregulated electricity
market was being proposed. As part of this change the state government was looking
at breaking up the generators to ensure they were competitive when they moved to the
market environment. This network evolution was an important step as the electricity
generators now would be treated as a business and would be required to return profit. Up
until then they had been solely government controlled and were treated as an essential
service. Being solely government owned their primary drivers were long-term availability
rather than profit. Once this breakup began Tarong Power Station ultimately became
part of Tarong Energy as a standalone business in the electricity market (Australia 2015).
Hence Tarong Energy wanted to better understand its business costs and part of this was
the need to understand its in-house electricity usage. Below is an outline of the level that
Tarong Energy and now Stanwell Corporation understand their internal energy usage. It
will start with an overview of the stations electrical distribution.
Figure 2.3 is a high level single line diagram showing the electrical distribution of a single
unit.
2.5 Energy Monitoring 17
Figure 2.3: Unit 1 Distribution Overview (Hitachi 2003).
It can be seen that the Unit 6.6 kV switchboards are fed from the step-down transformers
connected to the generator output. These are the A and B Unit Transformers and together
they supply all the major drives and auxiliarys for a single unit. The station ancillary
supplies are shown in figure 2.4. Some of the station supplies are coal handling, air
compression, administration, water treatment and ashing.
The four unit generator transformers feed into the grid and also supply their own in
house auxiliary energy requirements via 25 kV/6.6 kV step down transformers, the unit
transformers. The ancillary services common to all units are supplied by the station
transformers A and B which draw their power directly from the grid. The station trans-
former also can be linked to the unit 6.6kV switchboards to assist in unit start-up until
it can supply its own auxiliaries. Each of the four generator transformers and eight unit
transformers have an energy meter installed. It is these meters that are directly connected
to AEMO for market purposes. These meters have been updated routinely as part of the
stations network connection agreement. The then Tarong Energy implemented metering
on the major drives across one unit, Unit 4, and all primary feeds on the station switch-
boards. The approach would give them a good picture of site usage on which to base
future strategy. Some additional meters have been added to individual areas such as ash
and chemical plants loads. Administration building supplies had more metering installed
in recent years under the government reporting guidelines of the National Greenhouse En-
ergy Reporting Scheme (NGERS). Each segment of the energy metering was connected
2.6 Control System 18
Figure 2.4: Station Ancillary Distribution Overview (Hitachi 2003).
to the station Supervisory Control and Data Acquisition (SCADA) system via local Mo-
mentum Programmable Logic Controllers (PLCs). The local PLC polls the individual
meters to retrieve the blocks of relevant data from the energy meter via RS-485. This
data includes current, voltage, power factor and running totalisers. All data retrieved is
then read out of the PLC’s by the plant data historian.
2.6 Control System
This section outlines an overview of the Tarong Power Station site control system and
some of its features that will assist this project.
The control system is central to all industrial plants. In this way Tarong is no different.
The control system allows a high level of automatic responses to the raw signal inputs from
the field. It is this automated system that enables the plant operators to oversee and run
the plant. The control system and associated data historians allows engineers to monitor,
improve and optimise the plant. Tarong is now 30 years old and is operating past its
design life. The original Hitachi control system has been upgraded to the Siemens T3000
Distributed Control System (DCS). The majority of plant’s, station and units based, now
2.6 Control System 19
run on T3000 with a project currently running to migrate the remaining plants across to
T3000.
T3000 is Siemens flagship product that allows both logic and Human Machine Interface
(HMI) creation to occur whilst the plant processes are running. Operators can continue
to monitor and control the plant while engineering staff work on the system (Ltd 2015).
It is a powerful all in one product with hundreds of individual blocks available from which
to create screens, logic and sequences. T3000 can be easily altered, optimised or expanded
up to the resource limitation of each individual Central Processing Unit (CPU).
The unit plant energy usage is the focus of this project so the T3000 structure for one
unit will now be briefly described; this description is true for all units. The unit T3000
architecture consists of a central server and two major buses. The first bus is the automa-
tion bus which looks after communications between the Automation Processors (AP).
The second bus is the application bus which handles the communication from the plant
APs to the operator workstations and other services that requires plant data access. The
central server is a multiple slice high redundancy server that sits between the two buses
and runs the T3000 processes. There are eight APs on a unit application bus to handle
the large number of Inputs/Outputs (I/O) required to safely and affectively manage the
running the unit, around 7000 I/O in total. Each AP handles a specific area of plant to
reduce the level of inter AP communication required. For example AP 3 is Fuel and AP
1 is Turbine. There are separate high integrity I/O for Safety System control loops. An
additional virtual AP exists in the central server. In this virtual AP logic can be created
without consuming resources. At TPS the logic space that this virtual machine provides
is used to run plant performance calculation and other functions that are not directly
related to plant functionality.
Some of the I/O relate to the control and monitoring of drives. These drives are a variety
of valves, dampers, pumps and fans. The drive data that exists within T3000 is position,
status, current, voltage and power factor when available. The data from this I/O is ported
from T3000 through to the data historians. It is this drive data around auxiliary energy
consumption devices that will be utilised during this project.
2.7 Conclusion 20
2.7 Conclusion
In this chapter the background and operation of Tarong Power Station was explored.
Both a brief history and a plant overview builds a picture of the current plant status.
It demonstrates the need to understand the unit auxiliary energy consumption and the
control system tools available for use during the project. This overview of the energy
information assists in understanding the gaps that this projects outcomes aim to fill, and
the drivers behind them. The next chapter reviews the existing literature related to the
project aims.
Chapter 3
Literature Review
This chapter reviews relevant literature that is key to understanding previous work that
has been completed in the field of the project. The review will focus on the outcomes of
previous work to uncover any gaps and learning that can be utilised during this project.
The standards and best practice required in the field measurements of energy calculation
inputs will also be reviewed to assess Tarong Power Stations energy practices. Power
calculations, control optimisation and power station modeling techniques round out the
chapter.
3.1 Auxiliary Power Usage and Reduction
This section reviews the knowledge base of auxiliary power usage and the ways in which
auxiliary power reduction has been approached around the world. While there are some
common themes in the approach to energy reduction, each country has developed a unique
focus. Their focus is directed by the current issues that are seen to be the top contributors
to the problem.
The first steps in any attempt to reduce power usage is to understand the energy con-
sumption within a plant. This understanding often takes the form of an energy audit. A
Chinese study of energy auditing and its coal-fired power plant application by the Jilin
Province Science and Technology Department states that the concept of energy audit is
to inspect, examine and analyze the physical, financial and other high-energy using units.
3.1 Auxiliary Power Usage and Reduction 22
This study continues to state the targets of energy audit are investigating problems and
weaknesses in using energy, tapping of energy-saving potential, finding rectification mea-
sures and formulating saving goals (Li, Wang, Jiang & Zhang 2009). This project, whilst
not framed as an audit, will address the shortfalls in Tarong Power stations current lev-
els of energy usage knowledge. This current lack of knowledge is a primary barrier to
achieving meaningful energy reduction at TPS.
United States of America Network Experience
The 2011 Technical Report published by Electric Power Research Institute (EPRI) (EPRI
2011) focuses on the opportunities to enhance electric energy efficiency in the production
and delivery of electricity. The pertinent sections of the report are the chapters that
focus on electricity used for power plant auxiliaries. The report is centered around power
stations in the American grid. The discussions in this paper by EPRI are significant as
the American power industry, whilst considerably larger, is at a similar evolutionary state
as Australia. The report states that 11 % of electricity produced is used in delivering
the electricity to the customer of that 40 % is consumed in power production. The figure
3.1 shows some of the typically uses of auxiliaries or so called “parasitic loads” in ther-
mal power plants. EPRI states that “These auxiliaries are designed based on maximum
performance and environmental compliance, not on minimum use of in-house electricity
use.” (EPRI 2011). It is this statement which demonstrates that past design focus leaves
room for energy improvement which will be explored by this project. The paper discusses
that many auxiliaries have an opportunity to improve on their energy consumption due
to factors such as oversized auxiliaries, use of modulating mechanical mechanisms, design
trade offs between efficiency and cost, and tuning based on a particular parameter. It
is also noted that as is the case with Tarong the auxiliary load is monitored but not
generally considered separately as there are greater efficiency opportunities possible in
the boiler and turbine plant. EPRI consider plant auxiliaries as part of heat rate and
they contribute approximately 90.8 kHJ/kWh to the overall average heat rate of 1000
kHJ/kWh. This linkage may help lift the profile at Tarong. EPRI continue to say that
this amounts to 5-10 % of total generation depending on unit size, fuel type and ambient
temperature. This is a figure that is reflected in Tarong’s monthly reports. These reports
show that between 7.3 and 10.3 percent of generated energy is consumed by in house
auxiliary loads. These monthly figures are presented in table 3.1 (Sands & Blake 2015).
3.1 Auxiliary Power Usage and Reduction 23
Figure 3.1: Typical Auxiliary Loads in Coal Fired Power Plant (EPRI 2011).
Two facts that EPRI presented which are of particular relevance to the TPS situation.
The first was that the more base loaded and therefore higher capacity factor of the plant
the lower the internal usage. The second fact revealed by the figures is that age of the
plant is not as relevant as once thought. This is partly due to the additional of overall
auxiliary plant load by emission control equipment. The EPRI data is considered high
quality as it was collected over a 5 year period across 350 fossil fuel power stations.
World Comparison
The level of auxiliary energy consumption recorded in America by EPRI (EPRI 2011), of
5-10 %, is reinforced as an average level of consumption through the findings of Mandi
(Mandi et al. 2012) and India (Cornerstone 2014) in Indian power grid. Mandi reported
that within Indian sub-critical power stations, 30-500MW, an average auxiliary consump-
tion of 9.5 % existed. Mandi continued to state that this figure offers ample room for
improvement. This auxiliary energy usage level is also reflected in the recorded European
auxiliary energy consumption. ABB’s European productivity group states that between
7 and 15 % of coal-fired power generation is consumed in-house by auxiliary systems
(Boveri 2015). These figures reflect the levels recorded at Tarong Power Station in their
monthly reports (Sands & Blake 2015). Tarong’s Monthly reports over the last 12 months
show a range of 7.3 % to 10.3 % depending on the load profile of the month. Below in table
3.1 Auxiliary Power Usage and Reduction 24
Month Unit 1 Unit 2 Unit 3 Unit 4 Tarong Total
August 2015 7.7% offline 7.9% 8.0% 7.9%
July 2015 8.1% offline 7.9% 8.2% 8.1%
June 2015 8.8% offline 8.0% 8.3% 8.3%
May 2015 8.1% offline 7.3% 8.0% 7.8%
April 2015 8.1% offline 8.2% 8.2% 8.2%
March 2015 8.7% offline 8.7% 9.1% 8.8%
February 2015 8.6% offline 8.7% 8.9% 8.7%
January 2015 8.9% offline 8.9% 9.8% 9.2%
December 2014 9.0% offline 9.1% 9.3% 9.1%
November 2014 8.8% offline 10.3% 10.3% 9.7%
Table 3.1: Auxiliary Energy Usage Figures in Percentage by Unit for the Last 10 Months
(Sands & Blake 2015).
3.2 is a summary of the results uncovered from different countries. This identifies a key
point, which is that Tarong’s auxiliary energy consumption is consistent with recorded
levels of internal power usage in coal fired power stations around the world.
Location America India Europe TPS
Auxiliary Energy Usage Level 5-10 % 9.5 % 7-15 % 7.3 - 10.3 %
Table 3.2: A Sample of Auxiliary Energy Consumption figures from Around the World.
(EPRI 2011) (Mandi et al. 2012) (Boveri 2015) (Sands & Blake 2015)
Figure 3.2 shows the breakdown of the auxiliary power consumption for a typical coal fired
power plant (EPRI 2011). The EPRI figure shows that the BFP group is the primary
power consumer at approximately 37 % of the total load. The CW and ID fans make
up the next 30 %. This breakdown will be compared to the results found during this
project and will be used to drive the investigation portion of this project. The EPRI
report also presents some suggested methods for auxiliary power consumption reduction.
Its primary recommendation are the installation of adjustable speed drives (ASD) this
is also a recommendation made by ABB European Division (EPRI 2011) (Boveri 2015).
This is a major capital works investment but can, particularly at lower loads, deliver large
savings of up to 85 % reduction in drive power. The investigation of the application of
3.1 Auxiliary Power Usage and Reduction 25
Figure 3.2: Typical Auxiliary Loads and their Breakdown of Usage in Coal Fired Power Plant
(EPRI 2011).
this innovation to Tarong plant is outside of the scope of this project. Though it should
be noted that while drive replacement is not to be investigated by this project the idea
has merit and should be investigated by Tarong in the future. This project will instead
consider control optimisation of the high usage drives to reduce energy consumption.
Indian Network Experience
The concept of reducing the auxiliary power usage of coal fired power stations was explored
in Indian power stations by Rajasheker, Mandi, Udaykumar and Yaragatti. The focus
was how to reduce auxiliary power usage without major capital investment. Their study
was undertaken across twenty three 210 MW sub-bituminous coal fired power stations
giving a large base for data collection and comparisons (Mandi et al. 2012). The data
within the paper is closely representative of the Tarong units as the cycle, depicted in
figure 3.3, is very similar to the Tarong cycle previously outlined. This study is highly
relevant to the Tarong Power Station situation as Mandi discusses that the Indian power
stations are routinely requested by the network controller to run at sub-optimal partial
load. As discussed previously this is routinely the case for TPS as it is no longer always
3.1 Auxiliary Power Usage and Reduction 26
Figure 3.3: 210MW Coal Fired Power Plant Cycle (Mandi et al. 2012).
a base load station in the AEMO dispatch order of merit.
The major contributors to auxiliary power usage are the BFPs, IDs, FDs, PAs, Mills,
CW pumps and CEPs. Mandi’s work showed a strong link between plant load factor
and auxiliary power usage. Load factor is the actual running power level versus the
peak running power level expressed as a percentage. The work undertaken by Mandi
and associates show that when the plant is running at higher loads all major drives are
operating at a high load factors. This leads to the plant operating more efficiently and is
reflected in the overall reduction of auxiliary power usage (Mandi et al. 2012). The graph
in figure 3.4 is extracted from the study which shows the positive impact of load factor
on each major power contributor.
Mandi’s concludes that the auxiliary power consumption within the Indian power gener-
ation plant is on the high side compared to other developed countries and that low load
factor is a primary cause of running with higher auxiliary power usage than the benchmark
(Mandi et al. 2012). The Tarong monthly reports reflect this occurrence as when running
at higher unit loads, therefore higher load factor, TPS shows lower recorded auxiliary
power usage (Sands & Blake 2015). Mandi lists the following Indian specific factors that
affect load factor: Coal quality, lack of equipment optimisation, excess steam and water
flow, hesitation to upgrade equipment, as well as poor initial design with many oversized
drives. All of these factors, coupled with inefficient controls, have a degrading effect on
3.1 Auxiliary Power Usage and Reduction 27
Figure 3.4: Auxiliary Power Consumption Against Load Factor in Coal Fired Power Plant
(Mandi et al. 2012).
the plants overall power factor. There are many reasons presented for the lower power
factor of the units across India. Mandi gives reasons like inadequate coal supply, poor
coal quality, ID fan erosion, inadequate PA supply and non-availability of mills are all
related to the low level of development in the Indian power generation industry (Mandi
et al. 2012). These factors are largely not present within Tarong Power Station. If these
do occur i.e. intermittent lower coal quality, which does occur in the form of high iron or
ash content coal, it is well prepared for and effectively managed. During these lower coal
quality periods the plant does show downgrading effects. Some other factors they discuss
are still a large concern within the Tarong plant, however it should be mentioned that they
are more quickly and effectively addressed due to the developed nature of the Queensland
electricity industry. Factors such as poor electrostatic precipitator performance (ESP),
poor condenser vacuum, air leakage in the air heater and duct system, inadequate circu-
lating water temperature and less demand in the power grid all negatively influence the
efficiency of the Tarong Power Station and Stations worldwide. Many of these factors are
also identified in the 2011 EPRI report on American power generation as negative influ-
ences on auxiliary energy usage levels (EPRI 2011). This suggests that the case Mandi
puts forward of these being solely Indian specific factors is somewhat overstated.
3.1 Auxiliary Power Usage and Reduction 28
The Indian paper also presents some methods to improve plant load factor for each of the
major drive groups. Primarily Mandi focuses on health and capacity adequacy of major
auxiliaries, especially the Mills and ID fans (Mandi et al. 2012). This focus is interesting
and leaves out the major contributor to the auxiliary load and that is the BFP units.
The BFP influence and potential reduction will be explored more during this project.
The majority of energy conservation measures Mandi suggested to improve load factor
are already in place within Tarong to some extent or are not relevant due to local factors.
They are however relevant for discussion. In the paragraphs below each of the factors
discussed by Mandi are broken down by plant group. They are then reviewed for potential
relevance to this current project and Tarong in general.
Coal Pulveriser Improvement
Suggested mill group improvement factors, including improvements to coal quality, coal
grind sizing quality and raw coal sizing are all in place at Tarong. Coal quality is contin-
ually monitored but is limited by the available mine coal quality and contractual obliga-
tions. The station’s overall design is linked strongly to the station’s Meandu mine coal
characteristics. Trials of different coal quality have caused high heat pick up and ash build
up problems proving that TPS cannot cope with higher calorific value coal. The coal at
Tarong is sent through sizers to ensure that only coal under 25 mm diameter enters the
mills. As noted by Mandi the size of the coal entering the mill has a profound effect on
the mill group power consumption. The coal grind quality is monitored at Tarong and the
classifiers are a fixed design with regular inspections to ensure that the adjustment and
wear rates do not degrade the mill performance to greatly. Mandi suggests that hi-chrome
liners, wear plates and ball ring sets be utilised to reduce wear rates and improve mill
performance. This suggestion has merit within the broader mill management at Tarong
as wear rates and material review is continuous but is not in the scope of this paper. The
measure of periodic purging of mills to enhance the capacity and reduce auxiliary power
usage is a concept that has potential merit at Tarong and is in the scope of this project as
it can be fully implemented with control system modification (Mandi et al. 2012). This
being the case it will be fully investigated to evaluate the potential energy reduction value
in the Tarong environment.
3.1 Auxiliary Power Usage and Reduction 29
Induced Draft Fan Improvement
Most measures identified by Mandi to improve the ID fans capacity factor are strongly
linked to coal quality and plant overhauls (Mandi et al. 2012). As with the suggested
mill factors, most ID fan factors the Indian paper suggest are already in place at Tarong.
Reducing ash content in the coal, reducing pressure drops across the Precipitator and
Air Heaters, reducing furnace ingress and air ingress in flue gas ducts, improvement
of electrostatic precipitator performance and air heater performance are all in place on
Tarong units in some form (Hitachi 2003). Improving Ash content of coal through washing
the coal has always been a key part of the Tarong coal management strategy. It is achieved
through passing the coal through a wash plant on the mine site prior to being sent to the
Tarong stockpile. As noted by Mandi the ash content can be reduced through washing
which reinforces the Tarong practice. This reduction of ash content assists in the reduction
of pressure drops across precipitator and air heaters. Coupled with good soot blowing,
ashing and overhaul processes the pressure drop is addressed in these important high wear
portions of the process.
Another factor in ID fan energy usage is reducing furnace and air heater air ingress, tramp
air. This is managed at Tarong through O2 monitoring and the overhaul routines. This
is a major factor in ID fan performance and is largely out of the day to day control of the
operator. There is little that can be done from a control system perspective in relation to
tramp air. Mandi suggests that hard facing of the ID fan impellers to reduce the erosion
rate and enhance the capacity the ID fans (Mandi et al. 2012). This has merit within the
broader air management at Tarong as wear rates and material review is continuous but
is not in the scope of this project.
Another item raised by Mandi is the dosing of ammonia into the flue gas to assist the ESP
to collect particulate matter, this also has potential gains at Tarong but is not within the
scope of this project. There are no clear control system optimisation recommendations
for the ID fan group in the Indian paper (Mandi et al. 2012). This will be explored further
during this project as the ID fans are a significant contributor to auxiliary power load of
the coal fired stations.
3.1 Auxiliary Power Usage and Reduction 30
General Improvements
The remaining plant areas were viewed by Mandi under the section on other measures to
improve load factor. The improvements were very broad based wide reaching statements
that apply to all thermal plants but without any specific examples that could be directly
utilised during this project. Some improvements listed were improve boiler efficiency,
improve turbine efficiency, maintain circulating water temperature through cooling tower
maintenance, maintain regenerative heaters, use higher calorific value low ash coal, reduce
specific steam consumption and fuel consumption (Mandi et al. 2012). One improvement
suggestion that has control optimisation merit is to maintain generator performance by
enhancing cooling systems. This will be explored further during this project.
Whilst the majority of the auxiliary power reduction options explored by Mandi cannot
be realised without capital investment the concept of reduction through improving plant
load factor, operational optimization, adoption of advance control techniques and imple-
mentation of energy conservation techniques is certainly a target worth spending time
investigating as the repayment in considerable (Mandi et al. 2012). Most of the sugges-
tions within this key paper will not be pursued in the form raised. These suggestions will
be instead explored from the angle of capital free control modification.
Europe and South American Experience
ABB Europe suggests there are many ways that a reduction in auxiliary power can be
addressed. Primarily their focus is on installation of Medium Voltage (MV) variable speed
electrical drives. They claim a significant reduction in power consumption is possible if
the power required to drive the fans and pumps is generated in the right way. This
however requires significant capital outlay, whilst this may be repaid it is a major works
program to achieve (Boveri 2015). This has potential on the Tarong site but will not be
explored during this project.
In Brazil the focus of one research group into the reduction of auxiliary power in coal
fired plants was to investigate the influence of displacement power factor (DF) and total
harmonic distortion (THD) (Schwanz, Silva, Leborgne, Bretas & Gaidzinski 2014). The
paper concludes that the ESP section of the plant was the main source of THD and that
the major drive induction motors are the source of low power factor. They propose that
3.2 Power Calculations 31
these two factors combined are a source of energy consumption that could be addressed
through filters and capacitor banks. As Tarong has a similar pulsed DC precipitator and
multiple large induction motor loads it is a strong possibility that these two situations
exist within the TPS plant. This means that there is a potential to reduce power usage
at TPS with the capacitor and filter banks. This is outside of the scope of this project
and will not be explored further.
Auxiliary Power Usage and Reduction Conclusion
The literature discussed in this section has demonstrated that the auxiliary power is a
worldwide issue that causes the industry considerable problems each year. The EPRI
report suggests that if the auxiliary load in America could be reduced by just 5 % the
industry could delay the need for building new coal fired power plant by 10 years (EPRI
2011). There is clearly a need and incentive for this project and continuing research to
assist the industry move forward.
3.2 Power Calculations
This section outlines the theoretical background information that relates to power mea-
surement in a 3 phase AC system. Primarily all drives that relate directly to the auxiliary
energy consumption of the Tarong Power Station are 3 phase AC and generally are 6.6
kV or 415 V DOL drives.
When a load is connected to an AC power source it draws current. This current varies
sinusoidally depending on the load requirements and will lead or lag the voltage depending
on the load. The voltage supply system maintains the sinusoidal supply voltage that
originates at the generator. In a power generation environment, such as Tarong, it can be
noted that the voltage supply system is very robust being fed directly from the generator
output and a system of low impedance transformers. The other note is that loads within
the power generation environment tend to be inductive with a lagging current.
The information in this section is summarised from the text “Introductory Circuit Anal-
ysis” Chapters 14 and 19 (Boylestad 2003). There are 6 major concepts that need to
be comprehended in order to understand the power calculations. This understanding is
3.2 Power Calculations 32
required to enable the in-depth exploration of auxiliary power usage of a power station.
These are:
1. Average or Real Power;
2. Power Factor;
3. Apparent Power;
4. Reactive Power;
5. Power Triangle;
6. Wattmeters.
Each of these 6 will be mathematically explained in the sections below.
3.2.1 Average or Real Power
(Boylestad 2003) shows that Average or Real Power is defined as
p = vi (3.1)
Where
v = Vm sin (ωt+ θv) (3.2)
i = Im sin (ωt+ θi) (3.3)
Once manipulated using trigonometric identities the following represents the power de-
livered to a network.
p =
((VmIm)
2cos (θv − θi)
)−(
(VmIm)
2cos (2ωt+ θv + θi)
)(3.4)
It can be noted that the first portion of equation 3.4 is fixed i.e. not time varying and the
second portion varies with time t. This can be extended that the time varying portion
of equation 3.4, over a period of one AC cycle, will be equal to zero. Therefore the first
3.2 Power Calculations 33
portion of equation 3.4 is referred to as the average or real power delivered to the load.
This fixed portion becomes equation 3.5 below.
p =VmIm
2cos (θv − θi) (3.5)
As the difference between the angle of the voltage and the angle of the current is defined
as
θ = |θv − θi| (3.6)
Therefore the average power calculation becomes
P =VmIm
2cos θ (3.7)
The symbol for real or average power is P. Since the effective voltage and current equations
are
Veff =Vm√
2(3.8)
Ieff =Im√
2(3.9)
Therefore the average power calculation can also be written as
P = VeffIeff cos θ (3.10)
Where
P - is the average power in watts [W];
Veff - is the effective voltage applied to the load in volts [V];
Ieff - is the effective current drawn by the load in amperes [I];
cos θ - is the displacement angle between the voltage and current waveforms [degrees].
It is worth noting that if the load is purely resistive the voltage and current are in phase.
In this situation the phase angle is θ = 0 therefore cos θ = 1 and the average power is the
product of voltage and current. If the load is purely inductive the voltage and current are
in 90◦ out of phase. Since the i lags v by 90◦ the phase angle θ is −90◦ therefore cos θ = 0
and the average power is zero. If the load is purely capacitive the voltage and current are
90◦ out of phase. Since the i leads v by 90◦ the phase angle θ is +90◦ therefore cos θ = 0
3.2 Power Calculations 34
and the average power is zero. In situations where all components are not ideal the phase
angle will rarely reach +/− 90◦
Within Tarong it is the average or real power reading that is utilised to measure auxiliary
power consumption.
3.2.2 Power Factor
From 3.2.1 it can be seen that the displacement angle cos θ causes significant impact on
the final power value and it is therefore known as the power factor.
PowerFactor = PF = cos θ =P
VeffIeff(3.11)
Where
PF - is the power factor which is either leading or lagging [dimensionless];
θ - is the displacement angle between the voltage and current waveforms [degrees];
P - is the average power in watts [W];
Veff - is the effective voltage applied to the load in volts [V];
Ieff - is the effective current drawn by the load in amperes [I].
The terms leading and lagging power factor are used to describe if the current leads the
voltage or lags it. Capacitive loads have a leading power factor and an inductive loads
have a lagging power factor. As the auxiliary power consumption at Tarong is primarily
inductive loads it is expected the system power factor would be lagging.
3.2.3 Apparent Power
Is the result of the applied voltage and current without the influence of power factor.
This apparent power is generally used as equipment power rating in VA. The equation
for apparent power is shown below in equation 3.12.
S = V I (3.12)
3.2 Power Calculations 35
Where
S - is the apparent power [VA];
V - is the voltage applied to the load in volts [V];
I - is the current drawn by the load in amperes [I].
3.2.4 Reactive Power
The reactive power is the peak value of power that produces no net power transfer, either
in an inductive or capacitive load, during a complete cycle. The formula for reactive
power is shown in equation 3.13.
Q = V I sin θ (3.13)
Where
Q - is the reactive power [VAR];
V - is the voltage applied to the load in volts [V];
I - is the current drawn by the load in amperes [I];
θ - is the displacement angle between the voltage and current waveforms [degrees].
While the net value of this power is zero it is significant in that the load must be supplied
power every half cycle and return the power source during the other half cycle. This
reactive power calculation influences the peak level of power required to be supplied to
the network.
3.2.5 Power Triangle
Average power, Reactive power and Real power relate together in the vector domain as
in the equation 3.14 below
S = P +Q (3.14)
3.2 Power Calculations 36
Figure 3.5: The Power Triangle relates Apparent, Reactive and Average Power with Power
Factor (Boylestad 2003).
The power triangle is shown in figure 3.5. It visually displays the vector relationship
between the three powers and the power factor.
The individual power components of separate loads can be summated in the vector form
to give an overall resultant power vectors. If one quantity is not known then Pythagorass
theorem can be applied to the system to solve for the missing quantity. This is possible
as the real power and reactive power vectors always form a right angle.
S2 = P 2 +Q2 (3.15)
It is also noted that from the power triangle the power factor can be found and vice versa
as
PF = cos θ =P
S(3.16)
3.2.6 Wattmeters/Watthour Meters
A wattmeter is a type of meter that measures the number of watts consumed by a load in
a measure of time. The meter inputs are the voltage and current relating to the load. The
meter also measures the displacement angle between the two waveforms. It utilises these
three measurements to calculate the forms of power required to be displayed. Typically
the units of measure of this style of meter is kWhr or kilowatts consumed in an hour. The
meters normally totalise the kW consumed to allow billing or other monitoring to occur.
It is typical for the modern version of this style of meter to be connected to a remote
system to record and/or display consumption data.
3.3 Practical Industrial Power Measurement 37
The mathematical representation of this meter is
EWh = PW th (3.17)
Where
EWh - is the average or real power consumed by the load in watt hours [W/hr];
PW - is the instantaneous power consumed by the load in watts [W];
th - is the time in hours over which the measurement is taken [hr].
3.3 Practical Industrial Power Measurement
This section outlines the way in which power measurement is undertaken in industrial
plants and more specifically the way in which it is measured within Tarong Power Station
unit auxiliary plant.
From the late 1970’s until the turn of the century the primary means of achieving both
electrical drive protection and load metering was via a meter involving an analogue device
generally an induction disk type. These meters function by utilising the magnetic field
generated by passing a sample of the circuit load current through a coil which in turn
spins the induction disk. The field strength increases with increased current flow which
in turn spins the disk faster. In drive protection once the disk is moved far enough the
relay will trip the feeder. In energy measurement the speed of the disk can be used to
meter the loads energy consumption. Today the primary method installed in new plant
is digital based metering. These use the load voltage and current measurements together
with digital sampling times to calculate power flow. The meter than can process this data
in many ways to perform protection functions or measurement recording (IEEE 2013).
The key inputs to calculate power are voltage, current and power factor or the magnitude
of the phase angle θ. The measurement of these will now be discussed.
The first key measurement is the supply voltage to the load of interest. This voltage
measurement is difficult to measure directly in most cases as the supply voltage is high
voltage. It is generally too high for direct measurement or input into electronic compo-
nents. This being the case the voltage that is required to be measured is stepped down to
3.3 Practical Industrial Power Measurement 38
a safe level. In most cases the voltage is reduced to 110 V AC and is achieved by step down
transformers commonly known as a voltage transformer or VT. These transformers have a
primary winding that is connected to the higher level voltage and the secondary winding
is connected to the energy metering device or control system interface module. VT’s have
a fast response and high resistance to saturation with fuses to provide protection from
over voltage or failure events (IEEE 2013).
The next key measurement of interest is the load current. Similar to load voltage the value
of the current is normally too high for safe measurement without an interface device. The
interface device in the case of current is a current transformer or commonly known as a
CT. The CT is directly wrapped around the conductor of which the measurement is to
be made. This conductor can be a core of a supply cable or in higher current situations
it is common to be a busbar. This CT steps the high current value down to a lower value
that can be directly fed through an energy meter or control system interface module.
The final measurement is power factor and this is not measured as such, it is extracted
from the voltage and current measurement. It is the displacement angle between the two
measurements.
The accuracy of measurement is a combination of the accuracy, repeatability and linearity
of the pickup devices, namely CTs and VTs and the quality of the energy meter being
utilised to totalise the measurement. CT’s come in different classes, protection, metering
and general purpose. Protection class CT have a high saturation level as they are re-
quired to accurately measure high current under fault conditions. Metering CTs are high
accuracy and have good linearity in the region of interest. General purpose CTs have an
average level of accuracy across a wide operating range and are generally used for control
system feedbacks (IEEE 2013).
An article produced by Power Logic in conjunction with the University of Missouri-
Columbia describes the way in which energy metering can be combined with local Pro-
gramable Logic Controller (PLC) and computer based data historian to create a power
measurement network (Asumadu, Devaney, Wallis & Bond 1990). This is precisely the
way that energy metering data is collected at Tarong. Tarong Power Stations unit auxil-
iary power usage is recorded by two primary meters per unit, one on each of the two unit
transformer feeds into the unit 6.6 kV switchboards. It is these two meters that totalise
the auxiliary power usage on the unit. The meters are Power Logic model DM6200 which
3.4 Standards 39
get their voltage and current inputs from the metering class CTs and VTs mounted in the
unit switchboard 6.6 kV incoming cubicle. These power logic meters are then connected
to the plant data historian via Modbus over ethernet. The unit power meters are in good
working order and are regularly certified. It is the readings from these meters that are
used to report on the unit auxiliary usage each week, month and for annual reporting.
On Unit 4 additional metering devices are installed on each 6.6 kV drive and the major
415 V drives. These meters are Electrex branc and are connected together on a RS-
485 network whose data communications is controlled by a local PLC. This PLC is then
connected to the plant data historian via ethernet. The individual drive data network
has been left decay. The local meters are generally operational but the communications
network to retrieve data from the meters has failed. This has in the past been related to
individual meter failures dragging a segment of the RS-485 network down. Attempts to
repair this system are current underway but are unlikely to be completed by the end of
this project.
This section has summarised some of the key points of industrial power measurement.
3.4 Standards
This section outlines the standards relating to the measurement of energy metering.
The most relevant standard uncovered during the literature review was IEEE Standard
3001.8 - IEEE Recommended Practice for the Instrumentation and Metering of Industrial
and Commercial Power Systems (IEEE 2013). This standard covers Instrumentation
and Metering. It outlines the importance of metering including considerations to apply
the latest metering technology. It discusses examples of ideal installations, instruments,
permanent and portable meters, recording instruments an auxiliary devices such as CT’s
and VT’s. The standard provides guidance on the class of CT’ and VT’s for reliable
measurement.
As this project will utilise data from an existing metering network the standard has been
used to review the metering network and its overall capacity for accurate readings as
compared to the standards recommended practice. The TPS installation conforms well
to the standards outlined in IEEE 3001.8 - 2013. The majority of installation practices
3.5 Control Optimisation 40
are closely aligned with the example installations given in the standard. In particular the
class of CT being utilised is better than recommended, the VT’s limit the voltage and
have the relevant fuse protection recommended by the standard. The digital power meters
themselves are integrated into the industrial plant supervisory system as recommended
in the previous section, though current plant failures need to be addressed (IEEE 2013).
The device used for DCS feedback do not comply with the metering class standard outline
in IEEE standard. The devices accuracy combined with the 100 ms sampling time of the
control system makes them accurate to approximately 1 % meaning that they comply
with the standard expected accuracy under IEEE 3001.8.
As as result it is confirmed that the readings available in the TPS DCS are of a reasonable
standard from which to base calculations but will have a level of inherited error.
3.5 Control Optimisation
This section outlines the control optimisation and the recent research in the field.
Control optimisation is defined as the process of tuning a process to give an ideal response
in relation to a specific parameter. In the power station environment the units control
is optimised or tuned to give the optimum energy return for the coal being fed into the
boiler. This involves multiple loop tuning with the end aim being the lowest amount of coal
required to deliver the high pressure steam at the ideal temperature and pressure to the
turbine. During this tuning process the level of energy consumption is not monitored or
reviewed in any way. It is considered a byproduct of running the unit efficiently. Simon
and associates discuss in their 2010 control conference paper that primarily industrial
control and specifically power station control is based on multiple Single Input Single
Output(SISO). Often these loops are Proportional Integral Derivative (PID) that require
independent tuning even though they all influence the same final overall system output
(Simon et al. 2010). Simon’s focus was on adding in a overall controller to coordinate the
effort of the multiple SISO loops in achieving the steam temperature at minimum coal
input.
A simplified control scheme of a coal-fired power plant is shown in Figure 3.6. It visually
displays the primary inputs and outputs of the overall power station control system.
This simplified system shows the final drivers on which normal power station control
3.6 Power Station Modelling 41
Figure 3.6: The Simplified Control Scheme of a Coal-fired Power Plant (Simon et al. 2010).
optimisation is focused.
3.6 Power Station Modelling
This section outlines the styles of modelling recently utilised to represent power station
processes for the means of research or investigation. This project will require a robust
MATLAB model of auxiliary energy usage to be created. This model will be used to not
only better understand the energy consuming devices on the Tarong coal-fired units but
also to trial any possible reduction options that will be generated later in the project. The
intention of this section is to review the information known about power station modeling
techniques.
In 2011 a paper was presented to a Mechatronics seminar in China which gave an
overview of modelling and simulation of thermal power plant (Changliang, Hong, Jin-
liang & Chenggang 2011). This paper was reviewed as a starting point to understanding
the challenges of modelling power plant. The paper firstly reviews thermal process and
3.6 Power Station Modelling 42
its modelling. It acknowledges that whilst approaches are widespread from step-response
models to predictive control models there is always a large gap between the model out-
put and the reality of the plant process. This is primarily due to the complexity of the
thermal process. Zhen and associates next cover simulation of the thermal plant though
Distributed Control System (DCS) interface. These simulators use the actual DCS logic
and learned values from running the logic in the simulation environment to create a sim-
ulator that can be driven in real-time. Tarong Power Station has just such a simulator
for their 350 MW units based on the T3000 DCS platform. It is proposed as part of this
project to use this simulator to test in near real-time any proposed energy saving modifi-
cations to confirm the control system response. The paper discusses the fact that whilst
many solid modelling results have achieved there are many challenges to be resolved. The
primary issue is achieving a model that accurately reflects the real plant. Another is how
to best use the abundance of operational data that is stored by the plant historians. The
final challenge presented is how to best control access to the data to allow simulation
development to continue. The balance between open access to data and secure access still
needs to be addressed.
The creation of a non-linear model in Simulink for a 300 MW coal-fired unit is dis-
cussed in a paper presented to the IEEE Industrial Electronic Society Conference in 2004
(Changliang, Jizhen, Yuguang & Xiuzhang 2004). The basis of the model is to break
down the power production process into smaller systems in the complex S-plane envi-
ronment and then combine as a total model. The model is heavily reliant on parameter
adjustments to ensure the model closely reflects the running plant. It also required large
amounts of development to get a relatively close working model before parameter tuning
can begin. They concluded that much future work is required to develop usable models.
With a considerable amount of data tracked and stored in the modern era it is a nat-
ural progression that this data is utilised as the basis of models (Yang, Wang, Zhang
& Chen 2010). The Beijing University paper explores this potential by focusing on the
real-time data collected at 600 MW coal-fired units in China to produce a model that
focuses on determining energy-saving potential in the plant. They explain that as the
power generation process is complex and non-linear environment the creation of models
which accurately reflect the real plant is extremely difficult. This leads to the creation
of models that only focus on individual equipment items or on steady state process as
the dynamic nature of power generation process cannot be reflected accurately. Their
3.7 Conclusion 43
modelling efforts extract data from the plant historian and then apply rules to classify
and correlate the data to provide theoretical guidelines for energy conservation within
coal fired power generation. Chen and associates use the cleansed data through a Spatial-
temporal Distribution Model of Energy Consumption based on Support Vector Regression
algorithm. Their conclusions were that the process of data mining can support real-time
or historical review of power generation process under many possible algorithms setup to
reflect the user requirements. They propose that it could be linked to real-time plant data
to reflect the current status of the plant energy consumption. The modelling technique
based on data mining holds particular interest as the modern control system at TPS hold
vast quantities of data that can be readily accessed. The data is not only available in
realtime though data links to T3000 but historical data is able to be accessed, via the
plant data historian PIMS which holds the last 6 years data.
This section has reviewed a cross section of literature relating to potential modeling
techniques that the could be utilised in achieving this projects aims. The discussion
has highlighted the challenges of ensuring that the final model matches the actual plant
response. This was identified as being particulary challenging in the complex non-linear
power station environment with many modelling techniques requiring multiple parameter
tuning to achieve any decent level of accuracy. The key discovery of the literature review
on power station modelling was that the environment lends itself to models based on plant
data. This data based model technique has many advantages, the main one is the link
ensures that the model inputs are actual plant data not calculated or estimated. This
real data input basis means that when constructed properly the final model outputs have
strong ties to the actual plant.
3.7 Conclusion
In this chapter the literature related this project has been explored. The use and im-
portance of energy monitoring was reviewed with the main finding that monitoring and
reporting is a key step to reduction of energy usage.
It has been demonstrated that auxiliary energy consumption in power stations worldwide
is an issue which remains largely unresolved. Many of the solutions proposed require
capital works to achieve energy reduction. These solutions, as with all unproven capital
3.7 Conclusion 44
works programs, come with risks. The primary risks are that the outlay does not yield
the return expected, along with project cost and time over runs. Potential short and long
term plant issues associated with changes to an isolated section of the plant design also
pose a creditable risk. These reasons combine to mean most plant owners are reluctant to
take on large scale plant changes in the pursuit of energy reduction. This lends support
to undertaking projects such as this that look at ways to reduce energy consumption
without capital investment.
Energy metering calculations were covered as these will form the backbone of this projects
outcomes. The way in which practical industrial measurement is achieved was explored.
Confidence was gained that the energy input measurements made in the Tarong Power
plant are reliable. This is an important point as it is these DCS inputs that will be utilised
for both the creation of energy tracking logic in T3000 and to prove the MATLAB models.
The IEEE energy measurement standard was reviewed with respect to the installed equip-
ment and this showed that the installed measurement devices are suitable on which to
base calculations. The chapter was concluded with a discussion on the possible power
station modeling techniques. These were explored and the conclusion was that using data
mining as the model basis is not only robust but the ideal way to represent the complex
non-linear environment of power generation processes.
The next chapter will explore the project methodologies employed to achieve the project
aims.
Chapter 4
Methodology
This chapter will outline the methodology utilised to reach the project aims. It will begin
with a justification of the project methodology selection then move onto detail of how
each of the methods were achieved. Being a highly dynamic power station environment
the method chosen must be able to be compared to the installed energy meters to give
the confidence needed to propose long term plant change.
4.1 Selection
This section will outline the development of the project methodology. The methodology
was completed after analysing the literature review, the available plant data and the
required outputs from this project. The use of both MATLAB modelling and T3000
energy tracking logic was decided upon after investigation of other methodologies. This
method allows both a calculated and real world insight into the auxiliary power usage of
the power station. The dual approach will allow robust confirmation of the MATLAB
models as there will be a running total, on the actual plant, to compare it against.
The primary method chosen is the creation of plant data based MATLAB models that
will reflect the actual plant energy consumption throughout the model running period.
MATLAB models of the auxiliary energy consumption of the power station electricity
generation process will be created. This method will assist with both understanding the
high usage items and in testing of potential energy reduction options. MATLAB has been
4.1 Selection 46
selected as the software in which to develop the required models due to its flexibility and
common use in engineering environments. It is a proven software that is known to perform
this function well and is regularly referenced in power station modeling IEEE papers found
during the literature review (Yang et al. 2010). This will align with Objective 2 to “Create
a SIMULINK model, in MATLAB, of the energy profile of a single unit”. These models
will be valuable tools for achieving project Primary Aims 1 and 2 being “To understand
the auxiliary consumption within Tarong Power Stations 350 MW coal fired units” and
to “Identify potential control based energy reduction strategies”.
The approach of creating a purely mathematical based model in MATLAB was discounted
as not being linked strongly enough to the running plant to enable justification for change
in the control system on the actual plant. It also became obvious during the literature
review that the power generation process is extremely dynamic making the creation of an
accurate model very difficult. Most modeling paper outlined previously discussed concerns
that the end product was unlikely to reflect the actual plant. The process of creation of a
plant data based model is seen to be more robust and fit for purpose (Yang et al. 2010).
One of the secondary objectives of this project is to achieve real energy reduction on the
running plant and therefore the method chosen needed to reflect the requirement to be
able to present data to Stanwell management in support of any proposed change. Another
method that was considered was solely utilising the historical data in the plant historian
to create trends and graphs of the usage in certain situations. This method may have
been robust in giving some resultant data to work with and base potential improvement
options on but it lacks the ability to be able to test the impact of any proposed changes.
Therefore the data available will still be assessed as part of the project process as it
can assist in guiding the creation of potential energy reduction options. Data stored in
the plants historical databases, Process Information Management Solutions (PIMS) and
InSQL, will be extracted to explore the plants energy usage under different operational
conditions. The data extracted from the historian was used as an input into the MATLAB
models created.
The second primary method will be the creation of real time energy tracking logic within
T3000. This project methodology will utilise the real time plant current and voltage
measurements available in the T3000 DCS to calculate power usage on all drives related
to auxiliary power usage. The field measurements have been confirmed as suitable for use
in this method. The literature review confirmed that the measurement devices comply
4.1 Selection 47
with standard expected accuracy levels. The digital control system sampling rate and
CPU cycle rate whilst high will still have a low level of sampling error. This method
will align with Objective 3 - “Create energy tracking logic in the T3000 Digital Control
System”. The tracking logic will calculate and totalise the power usage of auxiliary power
devices with available feedback. This data can then be summated in a number of ways
to assist in investigation of usage under certain conditions. This will create valuable
data for achieving project Primary Aims 1 and 2 being “To understand the auxiliary
consumption within Tarong Power Stations 350 MW coal fired units” and to “Identify
potential control based energy reduction strategies”. Another driver for creation of the
T3000 tracking logic was creating something that can be used during this project and into
the future. This logic will progress toward the Secondary Aim 1 “To leave in place energy
consumption information that the station can utilised into the future”. Once the tracking
logic is created it can forever be used by the plant owners engineering staff to look at
short and long term energy usage of individual drives across all units. This tracking logic
can potentially assist in short term rectification of high power consumption events and
in long term planning to reduce the overall level of auxiliary energy consumption. There
is the potential to setup alarms based on this logic to warn the operator of the unit of a
high energy consumption event. This would allow operator assessment and response to
occur during the event rather than solely reviewing the data at the end of each week or
month.
The tracking logic and MATLAB modeling will be validated against the limited installed
energy metering which strengthens the method selection. The tracking logic will be
installed on Tarong Unit 4 initially and the MATLAB models will represent one units
auxiliary energy consumption. The tracking logic will be installed on the other 3 Units
once confirmed as correct and functional.
This MATLAB model and T3000 tracking logic findings will complement the comprehen-
sive literature review that focused on energy calculations, coal fired power plant auxiliary
energy usage, auxiliary power reduction techniques, relevant standards and control op-
timisation. The literature review had a broad approach to uncover as much relevant
information as practical.
Building on this the methodology selected is shown below
Primary Method 1 Extract the mechanical and electrical parameters from the data
4.2 Initial Data Gathering 48
historian e.g. vane positions, scoop positions, current, voltage and power factor
data;
Primary Method 2 Create MATLAB models based on the extracted data. The first
group of models summate the voltage and current data to ensure it matches the
monthly energy meter reported value. The second group of models will be able to
alter vane and scoop position to alter the energy consumption pattern;
Primary Method 3 Collate all potential energy reduction strategies and assess the
resulting strategies utilising the MATLAB model;
Primary Method 4 Create energy tracking logic in T3000 for all drives where field
measurements are available;
Primary Method 5 Compare T3000 energy tracking and MATLAB model results
against the installed energy metering;
Primary Method 6 Propose advantageous low risk changes to the operational plant;
Secondary Method 1 Test proposed changes to the operational plant on the unit
simulator;
Secondary Method 2 Implement proposed changes on the operational plant;
Secondary Method 3 Create HMI and reporting based on the T3000 tracking logic.
Note The primary methods are those that are required to complete the project pri-
mary objectives. The secondary methods are those that will achieve the projects
secondary aims if time and risk assessment permits.
The next sections will describe the process of implementation of the chosen methods.
Firstly a section discussing the mechanics of the initial data gathering exercise will be
presented. This then leads into discussions of the MATLAB model and T3000 logic
creation processes. These processes are laid out in order of completion to build a picture
of how the validity of the created outputs was confirmed.
4.2 Initial Data Gathering
The first stage of the design process comprised exploration and collection of data. This
required a multifaceted approach to locate any data that was loosely or specifically re-
4.2 Initial Data Gathering 49
lated to the auxiliary energy consumption at TPS. The approach was to first collate the
information then later sort for relevance and decide on whether inclusion was warranted.
The primary sources that where consulted were
1. Drawings - Electrical, Mechanical and Process Instrumentation(P&ID);
2. Operations and Maintenance Manuals - Boiler, Turbine and Generator;
3. Energy Meter Data - Manual, InSQL data and PLC code of the site energy metering;
4. T3000 - PIMS Data and I/O relating to energy usage;
5. Plant Performance Reports - Weekly and monthly editions.
Each of these areas will now be expanded on in the sections below. It should be noted
that each area was revisites as new information or path of enquiry became available.
4.2.1 Drawings
Mechanical, Electrical and P&ID drawings exist for the majority of TPS plant areas. The
majority of drawings were supplied with the construction contract. These drawings have
been progressively expanded and modified to reflect the current installation as changes
have occurred on the plant. Each drawing has a revision number or letter to track the
revisions as they occur. Majority of the drawings accessed have had minimal revision
and reviewing the revision detail blocks show that the changes have not changed the
original design intent. These drawings are stored in electronic form within a software
package called Projectwise, a Bentley drawing management package. A smaller selection
of electrical drawings are stored in hard copy form in drawing libraries across the site.
These two locations were searched to locate relevant drawings. The drawings located were
in two main categories, heavy power and instrumentation and control (I&C).
The unit auxiliary power is measured at the unit transformer level. The heavy power elec-
trical drawings of the power feeds radiating from these transformers were located. From
there drawing detailing of the size and type of loads was determined. Interconnections
between boards were identified for later use in models and tracking logic. A sample of
identified drawings is listed in Appendix G.
4.2 Initial Data Gathering 50
Below in table 4.1 all of the auxiliary loads that emanate from the unit 6.6 kV transformers
are listing including there key rating data.
Load Rating Duty Voltage Ampere Power Factor
A Primary Air Fan 990kW 1 of 2 6.6kV 103A 0.6-0.85
A Forced Draft Fan 2950kW 1 of 2 6.6kV 311A N/A
A Induced Draft Fan 4550kW 1 of 2 6.6kV 503A N/A
A Unit 415V Transformer 2000kVA 1 of 1 6.6kV 175A N/A
A Precipitator Transformer 2500kVA 1 of 2 6.6kV 220A N/A
A Condensate Extraction Pump 1100kW 1 of 2 6.6kV 116A N/A
A Circulating Water Pump 1350kW 1 of 2 6.6kV 158A 0.7-0.75
A Boiler Feed Pump 5650kW 1 of 3 6.6kV 550A 0.65-0.87
C Boiler Feed Pump 5650kW 1 of 3 6.6kV 550A 0.65-0.87
Bunkering Transformer 1000kVA 1 of 1 6.6kV 88A N/A
C Unit 415V Transformer 2000kVA 1 of 1 6.6kV 175A N/A
B Primary Air Fan 990kW 1 of 2 6.6kV 103A 0.6-0.85
B Forced Draft Fan 2950kW 1 of 2 6.6kV 311A N/A
B Induced Draft Fan 4550kW 1 of 2 6.6kV 503A N/A
B Unit 415V Transformer 2000kVA 1 of 1 6.6kV 175A N/A
B Precipitator Transformer 2500kVA 1 of 2 6.6kV 220A N/A
B Condensate Extraction Pump 1100kW 1 of 2 6.6kV 116A N/A
B Circulating Water Pump 1350kW 1 of 2 6.6kV 158A 0.7-0.75
B Boiler Feed Pump 5650kW 1 of 3 6.6kV 550A 0.65-0.87
Table 4.1: Auxiliary Load - Detail of all Loads that Contribute to the Auxiliary Power
Consumption of the Unit (Alstrom 1981). Note N/A (Not Available) Power Factors were
Unable to be Located.
The second part of the drawing review was a search for signals relating to the power
calculations. It was found the majority of drives had a current feedback but there was
limited voltage and power factor feedback. These I&C drawings show the T3000 loop
from the field device through to the analogue input card. The drawings located are also
listed in Appendix G.
4.2 Initial Data Gathering 51
4.2.2 Operation and Maintenance Manuals
When TPS was constructed the contract included the supply of multiple sets of oper-
ations and maintenance manuals. The manuals were supplied with each contract area.
The primary areas of interest for auxiliary power consumption are the Boiler, Turbine
and Generator Operations and Maintenance manuals. The operations sections of these
manuals describe to the plant operator the way the plant process works and how to drive
each of the individual components. The maintenance sections of the manuals detail the
individual plant devices spare parts, recommended maintenance schedules, fault finding
information, and the manufacturers literature at the time of installation. These manuals
have been updated and expanded as the plant has been modified or as the plant has been
made redundant. The manuals also contain the initial commissioning and testing results
for the plant drives.
Where available, manufacturer data on the major auxiliary energy consuming motors,
pumps and fans were gathered from the respective manuals. This was found to be limited
with data being unavailable for many drives. The operational descriptions of each of the
items identified from the drawings were also obtained and reviewed. These describe the
preferred operational limits and plant configurations for each system. In the vast majority
of cases this is still how the plant is operated. Discussions were held with operations staff
and the plant control logic was reviewed to confirm the plant function. This occurred
during data gathering with a view to understand the plant limitations and operational
practices to be considered during the creation of any proposed energy reduction options.
4.2.3 Energy Metering Data
A specific energy metering manual was located during the data gathering exercise. This
manual contains details of the installation of Electrix meters onto the Unit 4 and Station
6.6 kV switchboard feeders (Godsmark 2001). A full list of drives and CT sizes was
located. This is included in Appendix H along with some installation drawings. This
manual showed the detail of the RS485 and PLC network that was installed. The available
registers that were setup to be retrieved from the energy meters is shown, a sample of
this is included in Appendix H.
InSQL is the oldest plant data historian that exists on the Tarong site for SCADA/PLC
4.2 Initial Data Gathering 52
data collection. Using the detail found in the energy metering manual a review of available
data in InSQL was then carried out. It was hoped that this would yield recorded data
for the Unit 4 auxiliary energy drives. It was found that the Station 6.6 kV energy meter
installation was active with large amounts of data available. The Unit 4 system has
fallen into disrepair as the only figures being actively reported on are the high level unit
transformers which have there own communications network. The unit data has only
recorded null values for many years with old data being written over. This has meant
no meaningful data relating to the auxiliary energy consumption was able to be collected
from the InSQL system.
4.2.4 T3000
T3000 is the very broad title given to the DCS monitoring and controlling the majority
of the plant within TPS. The T3000 monitors and actively drives all of the loads that
consume auxiliary power on the units. Within the T3000 function block code there is an
archiving function that stores selected values at the rate of the solve of the block involved.
The archive data is stored on the T3000 application server. This storage is finite and only
three years available within the system at any point in time. Each month the oldest data
is archived off onto DVD. There also exists a plant historian PIMS which accesses and
logs points of data from within the T3000 code. This then allows long term retrieval and
viewing of plant data without needing to access historical DVD data.
During the drawing collection and analysis process I/O was identified that pertain to the
flow of energy on the plant. These are listed in Appendix G and are primarily voltages
and currents within the power system. These I/O were then investigated in T3000 to see
if they were archived. Any points archived are then read out and stored in PIMS. All
points identified in the drawing review were found to be archived at 200 ms in T3000.
The PIMS archive settings for these points set to record at 1 second intervals. This means
that rapid changes in T3000 will not be recorded in PIMS. This will give a low level of
sampling error in the PIMS data as it is being archived. It is worth noting that as data
is retrieved by the PIMS interface, interpolation of stored data will occur if the data
requested is not available at the point in time requested.
As it was now identified that the energy calculation data was available the next stage was
to locate those points within the process that effect the load on the auxiliary drives. These
4.2 Initial Data Gathering 53
include damper, vane and valve positions that affect the air and water flows within the unit
processes. While it it a highly dynamic process that is influenced by multiple functions
the points identified are the primary influence and for the purposes of this project are
suitable for MATLAB data input. The identified process points were investigated to
confirm the T3000 and PIMS archiving rates. All points identified in the process review
were found to be archived at 100 ms in T3000. The PIMS archive settings for these points
set to record at points were investigated to look at T3000 and PIMS archiving rates. All
points identified in the process review were found to be archived at 50 ms intervals.
Once all relevant PIMS points were identified 24 hour blocks of information were requested
of PIMS. These blocks were chosen to reflect the varying loads and situations that the
plant is routinely requested to perform. The blocks of data were requested from PIMS
with an interval of 1 second. This gave output excel files of 86402 rows of data which 2
rows of header data showing the PIMS tag name and the tag description. A sample of
this data is located in Appendix F.
4.2.5 Plant Performance Reports
Performance of the TPS is always under constant scrutiny as small changes in performance
can cause large variations in running costs. To effectively collate and disseminate the
plant performance data the responsible engineer creates a weekly report highlighting the
key performance indicators. This weekly report is a a short 2 page presentation of data
relating to capacity factor, availability, coal consumed, fuel oil consumption, generated
MWhrs, exported MWhrs (AEMO), auxiliary energy MWhrs, efficiency and fuel oil stock
levels. Graphs relating to water, fuel and emissions round out the report. The plant
performance engineer also prepares a far more detailed 35 page monthly report which
rolls up all of the details of the weekly reports and dissects them further. The auxiliary
energy usage is presented here as an overall figure per unit in MWhrs and in percentage
of generation.
The level of information contained in these reports is of use to the project aim of under-
standing the unit auxiliary energy consumption. Twelve months worth of these reports
has been collected and the key factors summarised for later reference. These summaries
are analysed in the results and data analysis chapter 5.
4.3 T3000 Energy Tracking Logic Creation 54
4.3 T3000 Energy Tracking Logic Creation
This section details the way in which the T3000 energy tracking logic was created. The
creation of energy tracking logic in T3000 was addressed in a staged approach to build
confidence in the accuracy of the logic before applying it to all loads of interest.
As discussed in the sections above it was identified that the majority of drives had realtime
current feedback available. There is limited voltage feedback into the DCS with realtime
voltages available at the 20 kV and 6.6 kV level but none at the 415 V level. The power
factor readings are very limited with only the 2 available. They are read out of the unit
transformer energy meters at a slow poll rate.
The first item that was built in T3000 was a set of test logic to implement equation 3.10,
real or average power. It is repeated here for convenient reference.
P = VeffIeff cos θ (4.1)
This formula was implemented using basic blocks from the T3000 library such as add,
multiply and divide blocks. This logic was then tested by simulating the input parameters
with known values. Applying these simulation values allowed a known current, voltage
and power factor to be input into the logic. This allowed the logic accuracy to be confirmed
against the result of manual calculations. This allowed testing of a variety of function
blocks to reduce the logic complexity to a minimum. This initial test logic is shown in
figure 4.1.
Once this test logic was confirmed as functional a page of logic was created for the A
Unit Transformer. This drive was chosen as it has a functional energy meter whose value
also exists in T3000. This allowed comparison logic to be created for logic confirmation.
The test logic was copied and the real plant 20 kV level power factor, voltage and current
inputs were connected to the blocks in place of the simulated test inputs. The realtime
power reading from the energy meter was then compared to the realtime calculated reading
generated from the logic. The initial logic instantaneous power matched the energy meter
realtime power reading. The same logic was then created using the voltage and current
readings on the 6.6 kV side of the A Unit Transformer. As there is no 6.6 kV level power
factor reading the same 20 kV power factor has been utilised. These two reading were
4.3 T3000 Energy Tracking Logic Creation 55
the compared to confirm accuracy. The accuracy of the calculated logic was found to be
within 1 % of the the 20 kV energy meter reading.
The next stage of the tracking logic creation was to expand on the realtime calculated logic
to include blocks for totalisation of the energy consumption. The T3000 integration blocks
were utilised as part of the totalisation function to implement equation 3.17, watthour
meter. This was then fed into a series of first in first out (FIFO) and addition blocks
to allow daily, weekly and monthly figures to be captured. Resets and progression of
these blocks required the connection of logic to a “Time” block that has realtime links
to the application server clock. Since this meter is one of the two meters utilised for
reporting there is a daily, weekly and monthly totalised figures to which the tracking
logic can be compared. The logic shown in figure 4.2 shows the final configuration of the
logic blocks required to achieve the MWhr totalisation. This stage of the project took
some weeks to achieve the required accuracy. This was primarily because the multiple
delays required between the FIFO blocks took many iterations to achieve the required
reset order. After each logic change the logic was tested using the simulation function but
a full 24 hour, midnight to midnight bock, was required to confirm the function. Often
the results reflected in testing by simulation were not reflected in the actual logic response
to the change of day. This was due to the speed of applying simulations is different to
the speed of input changes during realtime solve of the blocks.
The accuracy of this A Unit Transformer logic was proven to be within 2 % of the energy
meter readings for a period of two weeks. Then a page of logic was then created for the
B Unit Transformer at the 20 kV and 6.6 kV level. This was chosen to be the second
drive as it also has field energy metering that exists in T3000. Once B Unit Transformer
24 hour readings were proven logic pages were then created for each feed on the unit 6.6
kV switchboards.
The naming conventions of the current readings in the T3000 caused some confusion as to
which voltage level the current reading was being measured. The confusion was unable to
be clarified by drawings as the poor T3000 naming convention exists on the drawings also.
Physical inspection of the plant was possible in some cases and this assisted in identifying
the current measurement location. This situation mainly occurred on the current readings
on the transformer feeds as it was unclear as to which side of the transformer they were
being taken. The final calculated results guided the final current reading designation.
4.3 T3000 Energy Tracking Logic Creation 56
Figure 4.1: Initial Test Real Power Calculation Logic - Screen Capture from Tarong Power
Station DCS (Siemens 2015).
4.3 T3000 Energy Tracking Logic Creation 57
Figure 4.2: Unit 4 A Unit Transformer Energy Tracking Logic - Screen Capture from Tarong
Power Station DCS (Siemens 2015).
4.4 MATLAB Models Creation 58
At this point a new page of logic was created to compare the daily totals from the different
sources. The aim of this logic was to ensure that all the calculated data across the drives
is accurate. Once the data was initially summated a large discrepancy existed between
the final sum and the real energy metering reading. Two main reasons for this existed.
The first was that some current readings assumed to be taken at the 6.6 kV level were
actually measured at the 415 V level. This was resolved by replacing the 6.6 kV voltages
with fixed 415 V levels. The second was that no allowance for the switchboard bus ties
existed. To resolve the bus tie error, logic had to be added to account for the status of
the C 6.6 kV bus interconnection to either A or B 6.6 kV bus. The bus tie statuses have
been used to modify the addition of the C bus energy consumption to the correct bus
total. The logic shown below in figure 4.3 is the final configuration of this code for A bus
connected to C bus. Similar logic exists for the B bus. The logic compares the 24 hour
total from the energy meters with the calculated values at the 20 kV level. On the same
page all of the auxiliary loads that feed out of the 6.6 kV unit switchboards are summated
together and then compared with the unit transformer energy meter reading that feeds
them. The summation of the MWhr totals of each bus section agreed with the overall
supply energy meters with an error of under 2 %.
When this tracking logic was complete all energy consuming loads at the layer below the
official auxiliary energy meters were being monitored. This was a major milestone of
the project as data was now available to achieve the aim of understanding the auxiliary
energy usage on a TPS unit.
4.4 MATLAB Models Creation
This section details the way in which the MATLAB models were created. The end aim of
the MATLAB model creation process is to have a model that firstly, accurately represent
the auxiliary energy consumption of a running unit across a wide range of loads. Secondly
it must display the individual load contribution to enable full understanding of the energy
consumption and finally it must be able to be manipulated to test energy reduction options
later in the project.
The order in which the MATLAB models were created was to ensure that the accuracy
of the models was confirmed at every stage of development.
4.4 MATLAB Models Creation 59
Figure 4.3: Unit 4 A and C Energy Tracking Logic - Screen Capture from Tarong Power
Station DCS (Siemens 2015).
4.4 MATLAB Models Creation 60
4.4.1 Conceptional MATLAB Script Creation
The first stage of the MATLAB model creation process was to create a small script that
would firstly import the voltage, current and power factor data from the 24 hour PIMS
Excel export. The script, InitialAuxPower.m then calculates the daily auxiliary energy
total. In this case the 7th of July was used as both PIMS and T3000 energy tracking
data was available.
Once the data has been imported the script next manipulates the data into separate
vectors to allow power usage calculations to be performed. Array multiplication using
element-by-element multiplication is then used to implement equation 3.10, real or average
power. This calculation gives an output array, for each drive, that contains a second by
second energy consumption value. The script next totals each drive array to give a overall
energy consumption for the 24 hour period. These drive totals are then added together
to give an overall bus total. To prove the script accuracy the result was then compared
to the T3000 tracking logic total and the energy metering total for the 7th of July. The
full script and its output from the 7th of July input file are in Appendix B. The final
values from the July 7th script outputs and the actual plant energy meters are included
below for reference in table 4.2. The table shows that there is only a small error between
the created MATLAB script and the real plant energy meter the percentage difference is
shown in brackets to be at worst 1 %. This confirms the accuracy of the script for future
use in cross checking the accuracy of any future models created.
Plant Energy Meter Total Tracking Logic Total MATLAB Script Output
A Bus 239.04 234.10 (-2 %) 240.37 (+1 %)
B Bus 225.08 223.65 (-0.6 %) 226.39 (+0.5 %)
Aux Total 464.12 457.76 (-1 %) 466.76 (+0.6 %)
Table 4.2: Auxiliary Energy Usage Figures Comparison between Plant Energy Meter, Tracking
Logic and the MATLAB Script Output. Totals are in MW/hr and the Percentage Error in
Presented in Brackets.
The methodology selected was to utilise a data input based model in MATLAB. One
challenge was that 24 hour periods of running at a fixed load do not occur in the TPS
plant. TPS units are constantly tracking the AEMO dispatch requirements after the
bidding process is complete. Stanwell units are no longer the top of the dispatch order
4.4 MATLAB Models Creation 61
due to running cost and are more often at partial load rather than full load. Therefore
it was decided to take samples of data from the plant at different loads to determine the
state of the process at various loads. This data was then used as the basis to create 24
hour fixed load files.
The next step in the process was to then create multiple import files for the different load
profiles. Data was retrieved from PIMS for loads between 140 MW and 365 MW. A row
of data at 13 separate load points across the range were selected out of the data. These
rows of data were used to create Excel files of a 24 hour length at those 13 load points.
These files represent the voltage, current and power factor of the plant if it was to remain
locked at a fixed load for the 24 hour duration. It should be noted that when the PIMS
data was reviewed there were small variations in all parameters at the same load. This
is due to the dynamic nature of power production and in particular the continual small
variations in heat generated through the combustion process. That is no two kilograms
of coal consumed by the process are the same. The data around the load points was
reviewed and an average response was selected to be expanded out to a full 24 hour file.
These 13 load files were then fed through the initial import and manipulation script
discussed above. The outputs of the script at the 13 load points will later be used to
compare the to the output of the SIMULINK models to be created at the next stage of
the MATLAB creation process. This was essentially a baseline to use as a comparison
when the SIMULINK models were created. The results are presented below in table 4.3.
Plant 140MW 150MW 175MW 190MW 200MW 225MW
A Bus 193 196 210 211 216 223
B Bus 172 173 182 199 206 212
Aux Total 363 369 396 416 423 437
Plant 250MW 275MW 300MW 310MW 325MW 350MW 365MW
A Bus 237 253 269 271 282 297 303
B Bus 228 243 259 265 266 282 289
Aux Total 474 510 538 545 548 579 592
Table 4.3: MATLAB Script Output - Auxiliary Energy Usage Figures at the 13 Load Points.
The next stage was to create a MATLAB SIMULINK model of the auxiliary energy usage
process for a single unit.
4.4 MATLAB Models Creation 62
4.4.2 SIMULINK Auxiliary Power Model Creation
The SIMULINK creation process started with implementing equation 3.10, real or average
power for a single drive. This was achieved using the basic maths blocks in SIMULINK to
solve the power calculation. The result of this calculation gives an instantaneous power
value for the drive in watts. This instantaneous power value is then divide by a constant
1000000 to convert it from watts to megawatts. This instantaneous MW value is then
integrated up over the time of the model simulation, in this case 86401 seconds or 24 hours
1 second, midnight to midnight. The integrated figure is then divided by the number of
seconds in one hour to convert it to MWhrs. In an approach similar to the T3000 test
tracking logic creation the voltage, current and power factor inputs were set to fixed values
to allow testing to occur. Once the testing was successful test logic for a full drive was
completed. In this case the drive selected was A ID Fan. Shown in figure 4.4 below is the
initial test SIMULINK model of A ID Fan at 190 MW. The resulting SIMULINK energy
consumption for the 24 hour period was 52.49 MWhrs with the MATLAB script result
being 52.4854 MWhrs. This proved the concept in SIMULINK and gave a much better
visual display of both the instantaneous and total energy usage with a much reduced
model running time.
Next a lookup table was added to allow selection of the input current value based on a
one dimensional lookup table that cross references the generator load set point, in MW’s,
with the current of the drive. The lookup table contents were obtained from the PIMS
data Excel files. Figure 4.5 below shows the test SIMULINK model of A ID Fan at 190
MW utilising the lookup table to select the correct current required. The resulting energy
consumed was 52.49 MWhrs which proved the accuracy of the lookup table concept.
With the concept of utilising SIMULINK and lookup tables proven the next task was to
create a model of the full unit. The concept was to be expanded to have all voltages,
currents and power factors selected from lookup tables based on the required load set
point. All table data was extracted from the 13 load files created from PIMS data.
The full unit model was created based on the A 6.6kV bus being tied to the C 6.6 kV
bus. This was the state of the 6.6 kV boards when the data was exported from PIMS.
Creating code to select which bus is tied together was not created as it was assessed as
adding no value to achieving the overall project aims. The model was laid out in sections
for clarity and to avoid high levels of crossed lines or repetition in the model structure.
4.4 MATLAB Models Creation 63
Figure 4.4: Initial SIMULINK model of Induced Draft Fan - Basic Concept Version.
4.4 MATLAB Models Creation 64
Figure 4.5: Initial SIMULINK model of Induced Draft Fan - Lookup Table Concept Version.
4.4 MATLAB Models Creation 65
Figure 4.6: Auxiliary SIMULINK Model - Power Version of Full Unit - A Bus Common
Calculations.
A common power supply section was first setup for each bus section. Figure 4.6 is the
power supply calculations for A 6.6 kV bus. Note that the bus voltage and power factor
are selected from one dimensional lookup tables. The same code was used for B 6.6 kV
bus and both A and B 415 V bus. As discussed in previous sections the 415 V bus power
supply calculations require the use of a fixed value of 415 V as no field measurement is
available to create a lookup table.
Beneath each of the common power supply sections the drives that are fed from that
particular bus section were laid out in order. Each drive then takes the corrected bus
voltage feed from the common calculations section. This bus voltage is then multiplied
with the current output from the MW to current lookup table to give an instantaneous
energy consumption reading in watts. This is then converted and totalised to give MWhrs
in the same way as the trial model. All loads on the one bus section are then summated to
give an overall energy consumption for the bus. The two bus section energy consumption
figures are then added together to give a total MWhr usage for the 24 hour period at
that load set point. Data links to the workspace were added to allow results to be more
4.4 MATLAB Models Creation 66
easily accessed and reviewed. Below in Figure 4.7 is a section of the power model showing
a couple of the A bus load calculations and the load totals. The data for the MW to
current, voltage and power factor lookup tables was taken from the plant PIMS data
utilised during the concept script creation stage of the project. Each table spans the 140
MW and 365 MW load range.
Table 4.4 below contains part of the results data from the power model at 190 MW.
The result from the script at 190 MW has also been included. It shows that there is
only a small discrepancy, about 200 kW(0.04 %), between the load sum calculations from
the script and the load sum calculations from the model. This proves the accuracy of
the SIMULINK power based model as the script accuracy was previously proven against
the plant energy meter reading. There is a slightly larger discrepancy between the total
power and the supply transformer sum of about 6 MW(1.5 %). This discrepancy also
occurs in the T3000 tracking logic with the load summation being slightly different to
the transformer calculation and the transformer calculation being closest to the energy
metering measurement. The fact that the power factors are only measured at these
transformers leads to the comment the transformer calculation is more accurate. It is
surmised that the same error is apparent in both the T3000 logic and the model because
the same common power factor, raw current and voltage values are being used therefore
the error is very similar.
Plant Power Model Load Sum Script Load Sum Script Transformer Sum
A Bus 210.55 210.65 (0.05 %) 214.89 (2.02 %)
B Bus 199.22 199.28 (0.03 %) 201.51 (1.14 %)
Aux Total 409.77 409.93 (0.4 %) 416.39 (1.5 %)
Table 4.4: Auxiliary Energy Usage Figures Comparison between the MATLAB Script Output
and the Power Based Model Output at 190 MW. Totals are in MWhrs and the Percentage
Error in Presented in Brackets.
Below are some assumptions and additional notes that are relevant to the use of the power
model. The assumptions were required as not all drives were in service when the PIMS
data was captured, that is there were some drives with missing data. The additional
notes guide the user and outline the limits of the model. The full power model is located
in Appendix C.
4.4 MATLAB Models Creation 67
Figure 4.7: Auxiliary SIMULINK Model - Power Version of Full Unit - A Bus Load Calcula-
tions.
4.4 MATLAB Models Creation 68
Model Setup - Duty Pump Sections and Generator Load Set Point need to be setup
using file AuxModel Setup Results.m for a single load set point or AuxPowerMW-
Step.m to step through a range of loads;
Model Limits - The Load Range of the model is 140 MW to 356 MW, which reflects
the operational range of the plant;
Model Outputs - The Model Outputs the Total MW/hrs from each drive and overall
totals to the workspace as arrays;
CEPs - A CEP load table is a copy of B CEP as B was duty during data capture,
only one pump is in service at any time, both motor and pump arrangements are
identical;
BFPs - A BFP load table is an average of B BFP and C BFP as B-C were duty during
data capture, two of the three pumps are in service at any time, all three motor and
pump arrangements are identical;
415V Loads - No field 415 V level voltage measurement available in PIMS so a fixed
voltage used, 415 V;
CW Pumps - CW Pump switch inserted to allow isolation of 1 drive below a selected
MW level;
Pulverisers - Pulveriser loads are under the 415V transformer loads. There are 3 to 5
mills are in service across the load range.
The auxiliary power usage SIMULINK model based on the drive power consumption was
found, that while accurate, was limited in its application for investigating potential energy
reduction options. This was due to only one drive set being able to be freely manipulated,
this being the CW pumps. This could be achieved though modification of the generator
load at which the model utilises one or two pumps. The next step was seen to be the
creation of a model that included the position of the drive load elements such as vanes,
dampers, scoops and valves. The next section is an outline of the modifications to the
power model that were made to include both the load element positions and bias inputs
to allow load modification.
4.4 MATLAB Models Creation 69
4.4.3 SIMULINK Auxiliary Position Model Creation
Firstly all of the loads were assessed to see if a direct link could be made between them and
a variable load element. This assessment included if the manipulation of these elements
then allowed load to be shifted to another drive without disrupting the overall power
generation process. It did not assess any potential long term plant impacts from the load
shifting. The load assessment in shown below in table 4.5.
Plant Area Drive Element Plant Configuration Capable of Load Share
A FD Fan Outlet Vane 1/2 of 2 Yes
A ID Fan Inlet Vane 1/2 of 2 Yes
A PA Fan Outlet Vane 1/2 of 2 Yes
A CEP Outlet Valve 1 of 2 No
C BFP Scoop Thruster 2/3 of 3 Yes
A BFP Scoop Thruster 2/3 of 3 Yes
A CW Pump Outlet Valve 1/2 of 2 Yes
A Precip DC Pulse 2 of 2 No
Bunkering Multiple Motors 1 of 1 No
A 415V Bd Multiple Motors 1 of 1 No
C 415V Bd Multiple Motors 1 of 1 No
B 415V Bd Multiple Motors 1 of 1 No
B FD Fan Outlet Vane 1/2 of 2 Yes
B ID Fan Inlet Vane 1/2 of 2 Yes
B PA Fan Outlet Vane 1/2 of 2 Yes
B CEP Outlet Valve 1 of 2 No
B BFP Scoop Thruster 2 of 3 Yes
B CW Pump Outlet Valve 1/2 of 2 Yes
B Precip DC Pulse 2 of 2 No
Table 4.5: Assessment of Auxiliary Load Drive Elements, Plant Configuration and there
Capacity for Load Share.
Once it was established which drives could load share, time was then spent to find PIMS
data on the controlling element for those drives. These elements positions were found to
archived in both T3000 and PIMS. The PIMS data for the load elements was exported
4.4 MATLAB Models Creation 70
for the same periods of time that the load data was exported. From this data, look up
tables were built to allow selection of load element position at different loads. These look
up tables were added to the SIMULINK power based model in front of the existing MW
to current tables. This meant that for those elements that were assessed as being able to
load share the look up tables were then split into two parts. Firstly a MW to position
table and secondly a position to current table. This split allowed a point at which bias
inputs could added. The bias inputs were added in such a way that if load was removed
from one drive it is automatically added to the corresponding drive with whom it can load
share. In this way it was proposed that the overall motive load would remain the same.
An example is if the “A” FD fan load is increased by 10 % of vane position then the “B”
vane position is decreased by 10 %. Below in Figure 4.8 is a section of the position model
showing a couple of the drive load calculations with the split look up tables and the bias
inputs to allow load variation. The setup script was expanded to add the bias inputs
which allows the plant configuration to be modified before the running of the model.
The script also output the key figures for the plant loads to allow later interpretation.
The same assumptions and application notes from the power model apply to the position
model. The full position model is located in Appendix D.
The investigation of the auxiliary load manipulation through bias of load elements was
found to be accurate and assisted in building a clearer picture unit auxiliary load con-
sumption. The model was found to be limited in the sense that while load element position
was able to be manipulated it was not clear whether transferring the same position change
to another load element would give the same drive capacity to the other drive. To give an
example if the scoop position on one BFP was reduced by 10 % and the scoop position
of another BFP was increase by 10 % would the net feedwater flow required by the unit
remain the same? If all element responses were linear this would be the case but this is
unlikely with the known characteristics of the elements involved. It was decided to create
a SIMULINK process based model. The process of creating the process based model will
be presented in the next section.
4.4.4 SIMULINK Auxiliary Process Model Creation
Firstly, all of the loads previously identified as being able to load share were assessed to
see which process measurement would best reflect a change to the load element. The load
4.4 MATLAB Models Creation 71
Figure 4.8: Auxiliary SIMULINK Model - Position Version of Full Unit - A Bus Load Calcu-
lations.
4.4 MATLAB Models Creation 72
assessment in shown below in table 4.6 only includes those that can load share.
Plant Area Drive Element Process Measurement Capable of Load Share
A FD Fan Outlet Vane Air Flow Yes
A ID Fan Inlet Vane Furnace Pressure Yes
A PA Fan Outlet Vane Bus Pressure Yes
C BFP Scoop Thruster Suction Water Flow Yes
A BFP Scoop Thruster Suction Water Flow Yes
B FD Fan Outlet Vane Air Flow Yes
B ID Fan Inlet Vane Furnace Pressure Yes
B PA Fan Outlet Vane Bus Pressure Yes
B BFP Scoop Thruster Suction Water Flow Yes
Table 4.6: Auxiliary Loads and their Identified Process Measurement.
Once the loads were assessed and the process measurement identified, the data relating
to those processes was exported from PIMS. The data was then processed and points
were identified that reflected the state of the process across the load range of the model.
Specifically at the same 13 load points utilised in the power model. The SIMULINK
position based model was used as the basis of the creation of the SIMULINK process
based model. The position model was modified to replace the position data with process
data. This meant that the split look up tables then became MW to process and then
process to current. This then allowed the auxiliary power usage profile to be modified by
the bias blocks by sharing process flow. The model was then able to reflect the impact of
the load sharing on the running processes required by the unit to ensure load generation
was not interrupted. This was not possible with either the power or position models. The
same bias names were used so no modification of the setup script was required. The use of
the bias input then changed from 0-100% to the process scale for example the feedwater
flow 0-566 kilograms per second.
The same assumptions and application notes from the power and position model apply
to the process model with the relevant additional assumptions below. The additional
assumptions were required as not all the process measurements being modified exist in
the field as measurements. The full process model is located in Appendix E.
ID Fans - There is no available flow measurement to assist in process sharing of this
4.5 Conclusion 73
plant group, control is to furnace pressure but flow is the loading factor of the ID
fan. Hence position sharing was left in place for these drives;
PA Fans - There is no direct PA fan flow measurement to assist in process sharing of
this plant group, control is to PA bus pressure but flow is the loading factor of the
PA fan. Air flow readings where PA enters the mills are available and these are
combined in T3000 to give a total PA flow reading. This has been assumed as being
equally shared between the PA fans. Hence the total PA flow reading was halved
and used for the lookup table points.
This SIMULINK process based model was found to be robust model with which to explore
the possible energy reduction possibilities. Figure 4.9 shows a section of the process model
showing a couple of the drive load calculations with the split look up tables and the bias
inputs to allow load variation.
4.4.5 Model Comparison Data
Table 4.7 below shows the comparison of the results between the MATLAB script and
the three SIMULINK models.
Plant MATLAB
Script
Model
SIMULINK
Power
Model
SIMULINK
Position
Model
SIMULINK
Process
Model
A Bus 210.65 210.18 210.65 210.08
B Bus 199.29 199.22 198.45 205.08
Aux Total 409.93 409.77 408.63 415.15
Table 4.7: Auxiliary Energy Usage Figures Comparison between the Three SIMULINK Models
at 190MW. Totals are in MWhrs.
4.5 Conclusion
This chapter presented the detail of the methodology selected. The selection of the mul-
tifaceted approach of literature review, data based MATLAB modeling and T3000 logic
4.5 Conclusion 74
Figure 4.9: Auxiliary SIMULINK Model - Process Version of Full Unit - A Bus Load Calcu-
lations
4.5 Conclusion 75
creation approach was justified as being required due to the project outcomes secondary
aims to alter the running plant. Being a highly dynamic power station environment the
dual approach and comparison to the installed energy meters gives the confidence needed
to propose long term plant change. The steps taken to implement the methodology were
presented in the order undertaken to develop the understanding of the underlying pro-
cesses behind the high level methods. The next chapter will present the results of the
method outputs.
Chapter 5
Results and Data Analysis
This chapter presents in raw form the results from the T3000 tracking logic and the
MATLAB models.
This chapter is to analyse the data collected from all sources and evaluate for future use.
The validity of both the T3000 tracking logic and the MATLAB models will be explored
to assess their suitability for use in proposing and testing reduction options. A total cost
to the business and the environment will also be presented.
5.1 Performance Reports Data
This section presents the data collected from the weekly and monthly Tarong Power
Station performance reports for Unit 4. This unit has been presented as it was the first to
have T3000 tracking logic installed. The other units show very similar results. The data
shows the generated MWhrs of the unit along side the auxiliary MWhrs consumed and
the financial impact of the consumption. The average August Queensland pool price of
$55 per MWhr has been used for the financial impact value.(Zecevic & EEUAA 2015) The
environmental impact is also included to show the influence of the additional generation
required to run auxiliaries. If auxiliary usage was reduced then the additional generation
would not be required therefore reducing emissions and running costs. This additional
available generation could alternately be sold to the market. The Stanwell corporation
figure of 3.664 tonnes of CO2 per one tonne of coal consumed has been used as the
5.1 Performance Reports Data 77
environmental impact figure. (Sands & Blake 2015)
The data table below 5.1 below is an extract of the vital statistics from the monthly TPS
performance reports that relate to auxiliary energy consumption. An additional column
for the estimated financial cost of this auxiliary energy usage has been added. This table
then gives a complete snapshot of the financial and environmental impact of the auxiliary
energy consumption requirements of the Tarong Power Station Units.
Unit 4 Aux Usage Aux % Aux Cost Coal CO2
Sept 15 19071 MWhr 7.8 % $1,048,905 8915 t 32664 t
Aug 15 18499 MWhr 8.0 % $1,017,445 8679 t 31799 t
Jul 15 16586 MWhr 8.2 % $912,230 7883 t 28883 t
Jun 15 15839 MWhr 8.3 % $871,145 8083 t 29616 t
May 15 17632 MWhr 8.0 % $969,760 8242 t 30198 t
Apr 15 16197 MWhr 8.2 % $890,835 7318 t 26813 t
Mar 15 14474 MWhr 9.1 % $796,070 6792 t 24889 t
Feb 15 13265 MWhr 8.9 % $729,575 6234 t 22844 t
Jan 15 12993 MWhr 9.8 % $715,615 6425 t 23544 t
Dec 14 13847 MWhr 9.3 % $761,585 7162 t 26244 t
Nov 14 12022 MWhr 10.3 % $661,210 5917 t 21682 t
Table 5.1: Auxiliary Energy Usage Figures - Unit 4 for the Last 10 Months (Sands & Blake
2015)
This Unit 4 auxiliary energy usage figures shown in table 5.1 that the monthly cost both in
financial and environmental terms is significant and that even small reductions could give
tangible returns to the business. The auxiliary energy required for the month is given in
both MWhrs and as a percentage of total generation. The effect discussed in the literature
review chapter of overseas power station requiring less auxiliary energy, as a percentage,
when the plants are running at a higher load factor can be seen within these figures.
That is during months that the plant runs at a higher load profile the required auxiliary
energy, as a percentage of total generation, is reduced. For example in September 2015
the average load of the unit was 339 MW and the cost of running the auxiliaries was over
$1 million dollars but the percentage of generation was 7.8 %. Whereas during November
2014 the market was depressed with a lower load profile leading to an average unit load of
162 MW. During this month the percentage of generation required to run the auxiliaries
5.2 T3000 Daily Auxiliary Total Data 78
was 10.3 %. These figures show a clear link between the unit load profile and the unit
auxiliary energy usage efficiency.
5.2 T3000 Daily Auxiliary Total Data
This section presents a selection of the data collected from the T3000 tracking logic created
during this project. This data is then analysed.
The data table 5.3 at the end of this section is a selection of daily totals from the Unit 4
T3000 energy tracking logic created as part of this project. The dates selected represent
a cross section of the average generator load set points since the T3000 logic was made
active. The averages have been in the mid to high load range for the units. The table
also includes the real 24 hour totals from the plant installed energy meters, marked as
(Real), for comparison to the tracking figures.
The Unit 4 figures confirm that there is only a small difference between the calculated
and metered energy figures. The data shows that there a some plant areas contributions
that can be considered as negligible. The precipitator, bunkering and unit 415V boards,
which the mills are a part, only contribute relatively small amount of load to the overall
total. The PA fans and CEP contribute a medium amount of load to the overall total.
With the higher contributing loads being the CW pumps, BFP’s, FD fans and ID fans.
The load contributions of each plant area change over the load range of the generator but
the the four major auxiliary load consumers do not change.
This leads to the question “How does the load distribution of TPS’s auxiliary energy con-
sumption compare to other stations?” The load distribution of American power stations
was presented in the literature review chapter and will be used again now for comparison.
The figure 5.1 shows two pie charts, the first is the typical distribution of auxiliary power
consumption within American coal fired power plant (EPRI 2011). The second chart is
the average auxiliary load distribution at TPS. The load distribution shown here is based
on the average of results from the created T3000 energy tracking logics daily 24 hour
totals. The distribution shows that the four major contributors in America equates to
approximately 75 % of the total load and are the BFP, CW pumps, ID fans and PA fans.
When this is compared to the TPS result it can be seen that the four highest contributors
are the BFP’s, ID fans, CW pumps and FD fans. Together their contribution equates
5.2 T3000 Daily Auxiliary Total Data 79
to approximately 80 % of the total auxiliary load. Whilst there is some differences the
primary contributor in both results is the BFP group. With the CW pumps and ID fans
being the next largest contributors, all be it in the reverse order.
The pulverisers or mills contribute approximately 9-10 % of the total load and as such
the project created T3000 tracking logic to monitor and totalise the mills energy usage.
The data table 5.2 below is a selection of daily totals, for the 415 V pulverisers, from
the T3000 energy tracking logic. It can be seen that generally the load is shared evenly
between the mills that are in service.
Date 6/07/2015 8/07/2015 28/07/2015 6/08/2015 8/10/2015
A Pulv 11.67 6.39 6.78 6.89 6.44
B Pulv 0 6 6.66 6.56 6.43
C Pulv 9.22 6.1 0 0 5.42
D Pulv 8.76 5.61 6.4 6.41 5.89
E Pulv 3.25 6.17 6.6 6.57 6.24
F Pulv 0 0 0 0 0
Table 5.2: Auxiliary Energy Usage Figures in MWhrs - Pulverisers - Output from T3000
Energy Tracking Logic Daily Totals.
This section has presented an overview of the data collected from the Unit 4 T3000
energy tracking logic at TPS created during this project. This has shown that the major
contributors and the load distribution of auxiliary energy consumption is comparable to
other stations around the world. The data has also identified that only four plant areas,
or nine major drives, contribute 80 % of the total auxiliary load consumption.
5.2 T3000 Daily Auxiliary Total Data 80
Figure 5.1: A Comparison of the Typical Auxiliary Load Distribution in Coal Fired Power
Plant to the Average Tarong Auxiliary Load (EPRI 2011).
5.2 T3000 Daily Auxiliary Total Data 81
Date 6/07/2015 8/07/2015 28/07/2015 6/08/2015 8/10/2015
Ave Gen Load 244.25 255.71 289.50 290.88 325.79
Generated 5982 6137 6948 6981 7819
A Tx (Real) 241.38 244.48 256.18 256.93 285.38
A Gen TX 242.83 245.83 257.59 258.38 286.98
A Unit TX 236.65 239.64 250.93 251.67 278.15
A PA Fan 14.79 15 15.69 15.75 15.58
A FD Fan 22.42 22.57 25.1 25.02 30.1
A ID Fan 54.15 54.36 56.15 55.9 57.08
A CEP 0 0 20.11 20.21 21
A CW Pump 33.33 33.44 34 34 34.04
A BFP 0 0 90.76 91.62 102.81
A 415 V Bd 27.72 28.4 11.33 11.42 18.92
A Precip 0.99 1 1.12 1.13 1.14
C 415 V Bd 8.65 8.73 15.99 15.77 15.47
C BFP 74.49 75.91 89.03 0 0
Bunkering 0.11 0.12 0.14 0.16 0.12
B Tx (Real) 228.67 231.28 262.72 263.53 278.2
B Gen TX 230 232.59 264.16 265.08 279.64
B Unit TX 224.24 226.8 256.82 257.57 270.94
B PA Fan 14.63 14.86 15.59 15.45 15.24
B FD Fan 22.17 22.29 24.49 24.41 26.32
B ID Fan 54.22 54.49 56.04 55.54 57.53
B CEP 18.9 19.09 0 0 0
B CW Pump 34.3 34.41 34.84 34.8 34.72
B BFP 75.66 77.08 0 90.54 101.89
B 415 V Bd 7.17 7.32 19.94 19.39 18.7
B Precip 0.07 0.08 0.07 0.07 0.07
Table 5.3: Unit 4 Auxiliary Energy Usage Figures in MWhrs - All Plant Areas - Output from
T3000 Energy Tracking Logic Daily Totals.
5.3 MATLAB Model Output Data 82
5.3 MATLAB Model Output Data
This section presents the outputs from the MATLAB models created during this project.
The methodology section discussed the evolution of the models and proved the accuracy
of the overall auxiliary energy figure produced by the models against the plant installed
energy meters. Each of the models produced not only an overall auxiliary consumption
figure, at different generator load set points, but also the consumption figures for each of
the individual drives.
5.3.1 SIMULINK Auxiliary Power Model
Table 5.5 appears at the end of the section and is the output of the Power Model when
run through the auxiliary power MW step m script. This gives a baseline result for each
drive to use for future reference.
When these figures are combined into plant area totals the same trends in major energy
consumption identified in the T3000 energy tracking data becomes apparent in the model
output data also. Below in table 5.4 is the total energy consumption by plant area based
on the Power Model output data.
Gen Load 150 175 200 225 250 275 300 325 350
CEP Total 16.0 16.7 17.4 18.2 18.8 19.9 20.7 20.7 21.5
PA Total 25.1 28.7 26.1 28.0 29.5 30.2 34.0 33.7 34.1
Other Total 32.7 35.4 34.8 36.0 37.0 40.2 39.5 39.0 39.4
FD Total 36.3 37.8 38.1 40.0 43.8 46.5 51.4 56.3 62.8
CW Total 64.2 65.4 66.3 66.8 67.6 68.8 69.2 69.1 69.6
ID Total 97.3 101.2 103.4 106.4 109.7 110.8 114.3 115.2 117.3
BFP Total 93.7 103.8 127.7 131.3 150.1 171.4 189.2 198.5 217.6
Aux Total 368.1 391.8 421.8 435.2 465.2 496.6 527.5 548.0 578.4
Table 5.4: Auxiliary Energy Usage Figures in MWhrs - Plant Area Totals from SIMULINK
Power Model.
When these figures are presented graphically it can be seen that some loads do not increase
5.3 MATLAB Model Output Data 83
Figure 5.2: Tarong Power Station Baseline Auxiliary Energy Consumption - Power Model
Plant Totals.
much of the load range, the CW pumps are good example of a static load. Other loads
however increase considerably as the machine loads up, the BFPs are an example of this
type of load. The power model plant totals are presented in figure 5.2.
5.3 MATLAB Model Output Data 84
Gen Load 150 175 200 225 250 275 300 325 350
A FD Fan 18.3 19.1 19.0 20.0 22.0 23.2 25.9 28.3 31.5
A ID Fan 48.7 50.8 51.9 53.4 54.8 55.3 57.8 57.1 58.5
A PA Fan 12.7 14.4 13.1 14.1 14.9 15.1 17.0 16.8 17.0
A CEP 16.0 16.7 17.4 18.2 18.8 19.9 20.7 20.7 21.5
A BFP 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
C BFP 48.2 54.7 63.5 65.4 74.9 85.4 94.6 99.3 107.7
A CW Pmp 32.0 32.5 32.7 33.0 33.3 33.9 34.0 33.9 34.2
A Precip 1.0 1.5 0.8 0.9 0.8 1.1 1.3 1.1 1.0
Bunker 0.0 0.0 0.0 0.0 0.4 0.4 0.1 0.2 0.2
A 415V 15.8 17.2 26.6 27.7 27.6 30.2 28.8 30.4 30.7
C 415V 2.8 2.8 8.1 8.4 8.6 9.0 9.2 15.4 16.1
B FD Fan 18.0 18.7 19.0 20.0 21.9 23.2 25.5 28.0 31.3
B ID Fan 48.6 50.4 51.5 53.0 54.9 55.5 56.5 58.0 58.8
B PA Fan 12.4 14.3 12.9 13.9 14.7 15.1 17.0 16.9 17.1
B CEP 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
B BFP 45.4 49.2 64.1 65.9 75.1 86.0 94.6 99.2 109.9
B CW Pmp 32.3 32.9 33.6 33.8 34.3 34.9 35.1 35.2 35.4
B Precip 0.9 1.1 1.2 0.8 1.2 1.0 1.1 1.3 1.2
B 415V 15.0 15.7 6.2 6.5 7.1 7.6 8.1 6.2 6.3
Gen Total 3600 4200 4800 5400 6000 6600 7200 7800 8400
Table 5.5: Auxiliary Energy Usage Figures per Drive in MWhrs - Output from SIMULINK
Power Model.
5.3 MATLAB Model Output Data 85
5.3.2 SIMULINK Auxiliary Position Model
The output of the Position Model when run through the auxiliary power MW step m script
is presented in table 5.7 at the end of this section. This gives a baseline result output
from the Position Model. These drive level consumption figures match the outputs from
the Power model.
When these figures are combined into plant area totals the same trends in major energy
consumption identified in the T3000 energy tracking data becomes apparent in the model
output data also. Below in table 5.6 is the total energy consumption by plant area based
on the Position Model output data.
Gen Load 150 175 200 225 250 275 300 325 350
CEP Total 16.0 16.7 17.4 18.2 18.8 19.9 20.7 20.7 21.5
PA Total 25.1 28.7 26.1 28.0 29.5 30.2 34.0 33.7 34.1
Oth Total 32.7 35.4 34.8 36.0 37.0 40.2 39.5 39.0 39.4
FD Total 36.3 38.1 38.1 40.0 43.8 46.5 51.4 55.6 62.8
CW Total 64.2 65.4 66.3 66.8 67.6 68.8 69.2 69.1 69.6
ID Total 97.3 101.2 104.8 106.4 109.7 110.8 114.3 115.2 117.3
BFP Total 93.7 103.8 127.7 131.3 150.1 171.4 189.2 198.5 217.6
Aux Total 368.1 392.1 423.2 435.1 465.2 496.6 527.5 547.2 578.4
Table 5.6: Auxiliary Energy Usage Figures in MWhrs - Plant Area Totals from SIMULINK
Position Model.
Whilst the results of the position model are very similar they are presented here in an
alternate graphically form to examine them from another perspective. With the data
presented in this form it can be seen that the total auxiliary load consumed for a 24 hour
period, at different generator load set points, increases as the generator output increases.
It is however apparent that as the generator load increases the rate at which the auxiliary
power loads increases is reduced. An example is that as load increases from 150 MW or
200 MW the auxiliary power requirement increased by approximately 50 MWhrs. But
when the load increases from 300 MW to 350 MW the auxiliary load increase is only 44
MWhrs. The position model plant totals are presented in figure 5.3.
5.3 MATLAB Model Output Data 86
Figure 5.3: Tarong Power Station Baseline Auxiliary Energy Consumption - Position Model
Plant Totals.
5.3 MATLAB Model Output Data 87
Gen Load 150 175 200 225 250 275 300 325 350
A FD Fan 18.3 19.0 19.0 20.0 22.0 23.2 25.9 27.6 31.5
A ID Fan 48.7 50.8 52.5 53.4 54.8 55.3 57.8 57.1 58.5
A PA Fan 12.7 14.4 13.1 14.1 14.9 15.1 17.0 16.8 17.0
A CEP 16.0 16.7 17.4 18.2 18.8 19.9 20.7 20.7 21.5
A BFP 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
C BFP 48.3 54.6 63.5 65.4 74.9 85.4 94.6 99.3 107.7
A Precip 1.0 1.5 0.8 0.9 0.7 1.1 1.3 1.1 1.0
Bunker 0.0 0.0 0.0 0.0 0.4 0.4 0.1 0.2 0.1
A 415V 15.8 17.2 26.6 27.7 27.6 30.2 28.8 30.4 30.7
C 415V 2.8 2.8 8.1 8.4 8.6 9.0 9.2 15.4 16.1
B FD Fan 18.0 19.1 19.0 20.0 21.8 23.2 25.5 28.0 31.3
B ID Fan 48.6 50.4 52.3 53.0 54.9 55.5 56.5 58.0 58.8
B PA Fan 12.4 14.3 12.9 13.9 14.7 15.1 17.0 16.9 17.1
B CEP 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
B BFP 45.4 49.2 64.1 65.9 75.1 86.0 94.6 99.2 109.9
B CW Pmp 32.3 32.9 33.6 33.8 34.3 34.9 35.1 35.2 35.4
B Precip 0.9 1.1 1.2 0.8 1.2 1.0 1.1 1.3 1.2
B 415V 15.0 15.7 6.2 6.5 7.1 7.6 8.1 6.2 6.3
Gen Total 3600 4200 4800 5400 6000 6600 7200 7800 8400
Table 5.7: Auxiliary Energy Usage Figures per Drive in MWhrs - Output from SIMULINK
Position Model.
5.3 MATLAB Model Output Data 88
5.3.3 SIMULINK Auxiliary Process Model
The output from the Process model is presented in table 5.9 at the end of this section.
It contains the output of the Process Model when run through the auxiliary power MW
step m script. This gives a baseline result output from the Process Model. These drive
level consumption figures match the outputs from the Power and Position models.
When these figures are combined into plant area totals the same trends in major energy
consumption identified in the T3000 energy tracking data becomes apparent in the model
output data also. Below in table 5.8 is the total energy consumption by plant area based
on the Process Model output data.
Gen Load 150 175 200 225 250 275 300 325 350
FD Total 36.3 37.8 38.0 40.0 43.8 46.5 51.4 56.3 62.8
ID Total 97.3 101.2 104.8 106.4 109.7 110.8 114.3 115.2 117.3
PA Total 25.1 28.7 26.1 28.0 29.5 30.2 34.0 33.7 34.1
BFP Total 93.1 103.8 121.0 131.3 150.1 171.4 189.2 198.5 217.6
CEP Total 16.0 16.7 17.4 18.2 18.8 19.9 20.7 20.7 21.5
CW Total 64.2 65.4 66.3 66.8 67.6 68.8 69.2 69.1 69.6
Oth Total 32.7 35.4 34.8 36.0 37.0 40.2 39.5 39.0 39.4
A Total 195.6 209.7 233.8 241.2 256.1 273.5 289.5 303.2 318.3
B Total 172.0 182.1 182.7 193.9 209.1 223.1 238.0 244.8 260.1
Aux Total 367.5 391.8 416.5 435.1 465.2 496.6 527.5 548.0 578.4
Table 5.8: Auxiliary Energy Usage Figures in MWhrs - Plant Area Totals from SIMULINK
Process Model.
An third alternate presentation of the baseline results of the process model plant totals
are presented in figure 5.4. The steep increase of the BFP across the load range is a key
point. The other noticeable feature is the steady increase of the ID fan load in contrast
the to continually changing slope of the FD fan group.
5.3 MATLAB Model Output Data 89
Figure 5.4: Tarong Power Station Baseline Auxiliary Energy Consumption - Process Model
Plant Totals.
5.3 MATLAB Model Output Data 90
Gen Load 150 175 200 225 250 275 300 325 350
A FD Fan 18.3 19.1 19.0 20.0 22.0 23.2 25.9 28.3 31.5
A ID Fan 48.7 50.8 52.5 53.4 54.8 55.3 57.8 57.1 58.5
A PA Fan 12.7 14.4 13.1 14.1 14.9 15.1 17.0 16.8 17.0
A CEP 16.0 16.7 17.4 18.2 18.8 19.9 20.7 20.7 21.5
A BFP 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
C BFP 48.3 54.6 63.5 65.4 74.9 85.4 94.6 99.3 107.7
A CW Pmp 32.0 32.5 32.7 33.0 33.3 33.9 34.0 33.9 34.2
A Precip 1.0 1.5 0.8 0.9 0.7 1.1 1.3 1.1 1.0
Bunker 0.0 0.0 0.0 0.0 0.4 0.4 0.1 0.2 0.1
A 415V 15.8 17.2 26.6 27.7 27.6 30.2 28.8 30.4 30.7
C 415V 2.8 2.8 8.1 8.4 8.6 9.0 9.2 15.4 16.1
B FD Fan 18.0 18.6 19.0 20.0 21.8 23.2 25.5 28.0 31.3
B ID Fan 48.6 50.4 52.3 53.0 54.9 55.5 56.5 58.0 58.8
B PA Fan 12.4 14.3 12.9 13.9 14.7 15.1 17.0 16.9 17.1
B CEP 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
B BFP 44.9 49.2 57.5 65.9 75.1 86.0 94.6 99.2 109.9
B CW Pmp 32.3 32.9 33.6 33.8 34.3 34.9 35.1 35.2 35.4
B Precip 0.9 1.1 1.2 0.8 1.2 1.0 1.1 1.3 1.2
B 415V 15.0 15.7 6.2 6.5 7.1 7.6 8.1 6.2 6.3
Gen Total 3600 4200 4800 5400 6000 6600 7200 7800 8400
Table 5.9: Auxiliary Energy Usage Figures per Drive in MWhrs - Output from SIMULINK
Process Model.
5.4 Overall Auxiliary Energy Usage 91
5.4 Overall Auxiliary Energy Usage
This section is to discuss the overall auxiliary energy usage of Tarong Power Station and
present the baseline data against which future reduction options can be assessed.
The units auxiliary energy consumption at any single point in time can be represented
by a figure that is how many kWhrs auxiliary energy is required to produce 1 MWhr of
generator output. There is no single figure that can represent this value as the auxiliary
energy consumption efficiency improves as the generator load set point increases. The
baseline energy consumption has been calculated on the output of the process based
SIMULINK model. This auxiliary energy requirement is at a high of approximately 100
kWhrs per 1 MWhr generated at a generator load set point of 150 MW. The need for
auxiliary energy decays away to a low of approximately 70 kWhrs per 1 MWhr generated
at a generator load set point of 350 MW. This aligns with the discussion in earlier chapters
that the TPS units were design to be continuously running as a 350 MW base load station.
This baseline energy consumption pattern can be approximated by the mathematical
formula shown below.
y = 61 + 150e−x118 (5.1)
Where
y - is the average or real auxiliary power required by the unit to produce 1 MWhr of
power [kWhr];
x - is the generator load set point [MW].
When this equation is implemented across the generator load range the output plot is
shown in figure 5.5 being compared to the process model output across the same load
range. The result shows that the mathematical equation proposed can be utilised to
approximate the actual plant auxiliary energy needs.
5.4 Overall Auxiliary Energy Usage 92
Figure 5.5: Tarong Power Station Baseline Auxiliary Energy Consumption - Process Model
versus Formula.
5.5 Other Data 93
5.5 Other Data
This section discusses other data collected that requires exploration before the reduction
proposals are discussed in the next chapter.
Boiler feed pump loading is area that additional data has been collected. Being a major
auxiliary power contributor some additional knowledge of the pump system has been
collected. It was uncovered during this investigation that the pump curve and the actual
loading of the pump against this curve is monitored in realtime within T3000. A snap
shot of the A BFP pump curve, with the unit running at 350 MW is include in figure 5.6.
This shows that with the unit running at full load that the pumps are running at almost
there design limits. At this load that scoop position is approximately 50 % and the load
current is approximately 500 A. This current is approaching the maximum limit of 540
A found in the pump design data and T3000 alarms.
Some of load profiles from Unit 4 over the past year were collected to demonstrate the
range of loads that occur during the normal running of a unit. Each profile is a 24 hour
snapshot of the unit generator load output in MW. These are presented in figure 5.7 and
show that a daily unit load profile can vary considerably depending on market conditions,
time of year and Stanwell’s portfolio status. A unit can run for long periods at a set load
or if the market is particularly active it can change load continually throughout the day.
The first trend, on the top left, shows the unit running half the time at high load and
half the time at medium to low load. The second trend on the top right shows a high
load day with the load only coming down from 365 MW during the night and only then
to medium load of 220 MW. The bottom left trend shows a day when the market is over
supplied during a low state load period, the unit is at 140 MW all day. The last trend on
the bottom right shows a fairly depressed market day with the load being primarily 200
MW with periods at 140 MW.
5.5 Other Data 94
Figure 5.6: A Boiler Feed Pump Curve with the Pump Loading Shown in Realtime - Unit 4A
BFP 350MW (Siemens 2015).
5.5 Other Data 95
Figure 5.7: Load Profiles of Unit 4 Tarong Power Station. Top Left, 365MW-190MW. Top
Right, 365MW-220MW. Bottom Left, 140MW. Bottom Right, 140-200MW (Siemens 2015).
5.6 Data Shortcomings 96
5.6 Data Shortcomings
This section discusses any identified shortcomings in the data collected from all sources.
In the area of the pulverisers or mills the MATLAB models created did not go down to
the level of individual mill loading. The total mill consumption calculated from T3000
tracking logic is 5.8 % of the total auxiliary consumption. The mill loading was included
in the higher level measurement of the unit 415 V transformers within the MATLAB
models. These transformers supply a 415 V distribution board that supplies multiple
drives. The mill motors are fed from these boards. Time was not spent on drives below
the 6.6 kV level as the data has shown that 80 % of the energy consumption is contributed
by the four major 6.6 kV drive groups, BFP’s, ID fans, FD fans and CW pumps.
There was no PA flow data available at the output of the PA fans themselves. The vane
position controls to the PA bus duct pressure so the flow is not directly required. The
individual PA air flows into the mills are measured and these six figures are combined into
an overall PA air flow. The data within the process model look up tables that relates to
the PA fans air flow used this combined figure divided by two as the required flow data.
This has assumed that the total air flow is supplied equally by the 2 fans. For this project
the equal distribution of flow is seen to be fit for purpose and allows the sharing of air
flows to be explored.
There is no ID fan air flow readings available with the vane position controlling the furnace
pressure as a negative. No individual air flows leaving the boiler are available meaning
that the position sharing arrangement from the position model has been left in place for
the process model.
The model makes no specific allowance for the BFP leak off operation. The leak off valves
operate between the BFP outlet and the DA to ensure minimum flow through the BFP is
always achieved. As the models range starts at 140MW the leak off valves are normally
closed.
The data within the models in relation to the precipitator and bunkering boards is con-
sidered to be slightly flawed. These flaws have been allowed to stay as the total auxiliary
power consumption of these systems is less than 0.5 % of the total. The models look-up
tables could not easily be made to accurately reflect the actual load characteristics of
5.7 Conclusion 97
these process. This is primarily because of the irregular batch process nature of these
systems and the shared output of the bunkering routes. The bunkering routes 3 and 4 can
both feed Unit 3, Unit 4 and Tarong North boiler bunkers. These routes are only started,
on average, once per shift to top-up the boiler bunkers. This is further complicated by
the fact that route 4 may top-up Unit 3 and 4 but the auxiliary power consumption is
only registered on the Unit 4 switchboard from which it is fed. The precipitator load
sits at a low level of continuous consumption with irregular spikes of relatively higher
consumption. These spikes have been ignored for the simplicity of the model and the
overall negligible impact of them on the auxiliary usage.
5.7 Conclusion
In this chapter the collected data from all sources was presented. The data has shown
that the auxiliary energy consumption is a significant financial and environmental cost to
Stanwell corporation. The model results demonstrate the breakdown of this consumption
by plant area and identify the high level energy consumers. A mathematical relation was
established that reflect the baseline energy consumption patterns of the coal fired units
at Tarong. The chapter was rounded out with a brief discussion of data shortcomings.
The next chapter will examine potential reduction options and analyse their implications.
Chapter 6
Proposed Options of Energy
Reduction
This chapter is to present proposed options for energy reduction on the Tarong Power
Station 350 MW coal-fired generation units. A list of options for each plant area will be
presented along with an assessment via the MATLAB SIMULINK process model. The
options created will look solely at the energy reduction potential. Whilst the options
presented will be possible from a process requirement, the initial sections here will not
assess the potential long term impacts to the plant. Then a risk comment will broadly
discuss potential impacts.
The final section of the chapter on optimum solution will present a possible unit wide
low, medium, and high risk approach solution. The low risk will have minimal perceived
long term impact whilst the high risk proposal will have an perceived higher impact.
Each section will firstly outline the current operational method of the plant area and
its current auxiliary power consumption profile. The sections will end with a proposed
operational strategy and the energy impact as assessed by the MATLAB SIMULINK
process model result.
6.1 Induced Draft Fans 99
6.1 Induced Draft Fans
This section is to look at potential ways in which possible energy reduction in the area of
ID Fans could be achieved.
The system consists of two fans that draw air from the boiler through the precipitator
passes and discharge the air through the chimney flu. Each fan is driven by a DOL motor
with the inlet of fan controlled by a damper vane. Currently the control system runs with
both fans in-service across the full load range. The figure 6.1 shows the configuration of
the unit Air and Gas systems at full load.
Below in table 6.1 is the current usage summary for the Induced Draft Fan system running
at different generator load set points for 24 hours.
Gen Load 150 175 200 225 250 275 300 325 350
ID Sys 97.3 101.2 104.8 106.4 109.7 110.8 114.3 115.2 117.3
Table 6.1: MATLAB Process Model Output - Induced Draft Fan System Energy Usage Figures
at 9 Load Points to be used as a Baseline. Generator Load Set Point in MW and Energy
Totals in MWhrs.
Below are the possible energy reduction strategies considered. Each option has three
sections, the strategy, the auxiliary MW result, and a risk comment.
Reduction Option 1
Strategy
This option is to lift the furnace pressure set point to reduce the differential pressure
across the ID fans.
Resultant Usage
This option is seen to viable since the Control System refit has given tighter control over
the furnace pressure. The influence of this strategy was unable to be tested with the
location of the fan running within the fan curve unknown. It is highly likely to have
a measurable positive impact on the energy consumption of this plant area warranting
future exploration and testing. For the purposes of this testing an assumption of 10
6.1 Induced Draft Fans 100
Figure 6.1: Air and Gas System Components - Tarong Power Station (Siemens 2015).
6.1 Induced Draft Fans 101
% reduction in ID fan energy consumption has been made. This figure is in line with
discussions with mechanical engineering staff regarding the potential influence of lifting
the furnace pressure set point. Below in table 6.2 is the resultant usage summary for the
Induced Draft Fan group after reduction option 1 is implemented.
Gen Load 150 175 200 225 250 275 300 325 350
ID Sys 87.6 91.1 94.3 95.8 98.8 99.7 102.9 103.65 105.6
Table 6.2: MATLAB Process Model Output - Induced Draft Fan System Energy Usage Figures
at 9 Load Points after Reduction Option 1 Implemented. Generator Load Set Point in MW
and Energy Totals in MWhrs.
Risk Comment
The primary risk is that during a furnace pressure excursion that the boiler would be more
likely to trip rather than ride through as the initial margin to the trip point is already
reduced. Additional logic to cope with this possibility and the risk associated with its
implementation would need to be considered for this option to become a reality.
Reduction Option 2
Strategy
At lower loads run with only one fan calling the second into service only once the process
requires more air than one fan can supply.
Resultant Usage
Below in table 6.3 is the resultant usage summary for the Induced Draft Fan group after
reduction option 2 is implemented. This strategy achieves a 45 % reduction to the ID
fan group at 150 MW. This strategy is only effective at low loads as the fan supply
limit is reached somewhere between 150 and 175 MW. This effective limit is reinforced by
anecdotal evidence from the unit operators experiences with only one fan group in-service.
Risk Comment
The primary risk is that the second fan will not start when required and delay the unit
load up. This is considered a low risk. The increased starts on the fan, motor and switch
6.1 Induced Draft Fans 102
Gen Load 150 175 200 225 250 275 300 325 350
ID Sys 53.6 101.2 104.8 106.4 109.7 110.8 114.3 115.2 117.3
Table 6.3: MATLAB Process Model Output - Induced Draft Fan System Energy Usage Figures
at 9 Load Points after Reduction Option 2 Implemented. Generator Load Set Point in MW
and Energy Totals in MWhrs.
gear will cause increased maintenance concerned compared to the current running profile.
Reduction Option 3
Strategy
This is an extension of strategy 2 with only 1 fan in-service until 175 MW. Once the
second fan is in service the bias on the ID fan loading is setup to ensure that one fan is
always optimally loaded with the second varying as required by the process.
Resultant Usage
The most efficient point of the ID fan was unable to be identified from the documentation
located. The most efficient point was assumed to be the load at 350 MW. So for the
testing of this strategy the B ID fan was locked into the full load position as early as
practical, at 275 MW. It was unable to be locked in sooner due to the minimum position
set point of A ID fan. Table 6.4 below displays the resultant usage summary for the
Induced Draft Fan group under this strategy. This results in a 5 % reduction of energy
consumption in this plant area and an overall reduction of 1.1 %.
Gen Load 150 175 200 225 250 275 300 325 350
ID Sys 53.6 101.3 103.8 105.2 108.2 109.5 113.1 116.6 117.3
Table 6.4: MATLAB Process Model Output - Induced Draft Fan System Energy Usage Figures
at 9 Load Points after Reduction Option 3 Implemented. Generator Load Set Point in MW
and Energy Totals in MWhrs.
Risk Comment
The primary risk is that running with unbalanced air flows will cause uneven wear within
the unit airflow paths over long periods of time. The secondary risk is that the second
6.2 Forced Draft Fans 103
fan will not start when required and delay the unit load up. This is considered a low
risk. The increased number starts on the fan, motor and switch gear will cause increased
maintenance concern when compared to the current running profile.
6.2 Forced Draft Fans
This section is to look at potential ways in which possible energy reduction in the area of
FD Fans could be achieved.
The system consists of two fans that draw air from the atmosphere and discharge it
through the air heater into the boiler. Each fan is driven by a DOL motor with the outlet
of fan controlled by a damper vane. Currently the control system runs with both fans
in-service across the full load range.
Below in table 6.5 is the current usage summary for the Forced Draft Fan system running
at different generator load set points for 24 hours.
Gen Load 150 175 200 225 250 275 300 325 350
FD Sys 36.3 37.8 38.0 40.0 43.9 46.5 51.4 56.3 62.8
Table 6.5: MATLAB Process Model Output - Forced Draft Fan System Energy Usage Figures
at 9 Load Points to be used as a Baseline. Generator Load Set Point in MW and Energy
Totals in MWhrs.
Below are the possible energy reduction strategies. Each option has three sections, the
strategy, the auxiliary MW result, and a risk comment.
Reduction Option 1
Strategy
This option is to run at lower loads with only one fan. The second fan is called into
service only once the process requires more air than one fan can supply.
Resultant Usage
The resulting effect of this option on the FD fan groups energy consumption is an 13.5
6.2 Forced Draft Fans 104
% reduction at 150 MW. The second fan will still be called into service before the load
reaches 175 MW giving an overall 1.3 % reduction in unit auxiliary energy consumption.
Below in table 6.6 is the current usage summary for the Forced Draft Fan system.
Gen Load 150 175 200 225 250 275 300 325 350
FD Sys 31.4 37.8 38.0 40.0 43.9 46.5 51.4 56.3 62.8
Table 6.6: MATLAB Process Model Output - Forced Draft Fan System Energy Usage Figures
at 9 Load Points after Reduction Option 1. Generator Load Set Point in MW and Energy
Totals in MWhrs.
Risk Comment
The primary risk implementing this option is that the imbalance in airflow will cause
inconsistent wear and temperature imbalance within the boiler. There is a low risk that
the second fan will not start when called and delay the unit load up. Long term the
additional starts on the fans and switch gear holds and increase maintenance risk as
compared to the current running arrangements.
Reduction Option 2
Strategy
The second reduction option for this fan group is an extension of the first to initially
run with one fan below 175 MW. Once the second fan starts one fan is optimally loaded
through the remaining load range with the other fan modulating to deliver the required
air flow.
Resultant Usage
This strategy gives a 13.5 % reduction at 150 MW but overall leads to a 1.8 % increase
in the overall FD fan group consumption across the full load range. This amounts to an
overall increase of 0.18 % in the auxiliary energy usage of the unit.
Below in table 6.7 is the current usage summary for the Forced Draft Fan system.
Risk Comment
6.3 Primary Air Fans 105
Gen Load 150 175 200 225 250 275 300 325 350
FD Sys 31.4 37.9 38.7 41.5 46.4 50.8 52.6 58.4 62.8
Table 6.7: MATLAB Process Model Output - Forced Draft Fan System Energy Usage Figures
at 9 Load Points after Reduction Option 2. Generator Load Set Point in MW and Energy
Totals in MWhrs.
The primary risk is that the unit load up will be delayed if the second fan does not start
when called. The overall impact of this strategy is an increase energy usage.
6.3 Primary Air Fans
This section is to look at potential ways in which possible energy reduction in the area of
PA Fans could be achieved.
The system consists of two fans that draw air from the outlet of the FD fan and discharge
part of it through the air heater into the mills as hot PA. Some of the discharge bypasses
the air heater and is directed into the mills as cold PA. Each fan is driven by a DOL
motor with the outlet of fan controlled by a damper vane. Currently the control system
runs with both fans in-service across the full load range.
Below in table 6.8 is the current usage summary for the Primary Air Fan system running
at different generator load set points for 24 hours.
Gen Load 150 175 200 225 250 275 300 325 350
PA Sys 25.1 28.7 26.1 28.0 29.5 30.2 34.0 33.7 34.1
Table 6.8: MATLAB Process Model Output - Primary Air Fan System Energy Usage Figures
at 9 Load Points to be used as a Baseline. Generator Load Set Point in MW and Energy
Totals in MWhrs.
Below are the possible energy reduction strategies. Each option has three sections, the
strategy, the auxiliary MW result, and a risk comment.
Reduction Option 1
6.3 Primary Air Fans 106
Strategy
The first reduction strategy is to run with one fan in service until the process requires
the second fan.
Resultant Usage
The option give a 36 % reduction of the PA fan group power consumption at 150 MW.
This leads to a 0.22 % overall reduction in auxiliary energy consumption across the load
range. The second fan will be required at 175 MW.
Below in table 6.9 is the resultant usage summary for the Primary Air system
Gen Load 150 175 200 225 250 275 300 325 350
PA Sys 16.0 28.7 26.1 28.0 29.5 30.2 34.0 33.7 34.1
Table 6.9: MATLAB Process Model Output - Primary Air Fan System Energy Usage Figures
at 9 Load Points after Implementation of Option 1. Generator Load Set Point in MW and
Energy Totals in MWhrs.
Risk Comment
The primary risk is that the second fan will not start was required delaying the unit load
up.
Reduction Option 2
Strategy
This option is an extension of option 1 with one fan being run until the process calls
for the second fan. Once the second fan is called one is run at optimum with the other
modulating to match the process flow requirements.
Resultant Usage
This strategy gives a 36 % reduction of the energy consumption of the PA fan group at
150 MW. Over the full load range of the unit the resultant usage of the strategy is a 7
% increase of the PA fan group leading to a 0.45 % overall increase in auxiliary energy
consumption overall.
6.4 Circulating Water Pumps 107
Below in table 6.10 is the resultant usage summary for the Primary Air system
Gen Load 150 175 200 225 250 275 300 325 350
PA Sys 16.0 26.0 27.2 27.8 31.7 35.9 38.9 41.2 44.0
Table 6.10: MATLAB Process Model Output - Primary Air Fan System Energy Usage Figures
at 9 Load Points after Implementation of Option 2. Generator Load Set Point in MW and
Energy Totals in MWhrs.
Risk Comment
The primary risk is that the unit load up will be delayed if the second fan does not start
when called. The overall impact of this strategy is an increase energy usage.
6.4 Circulating Water Pumps
This section is to look at potential ways in which possible energy reduction in the area of
CW Pumps could be achieved.
This system consists of two pumps which have an hydraulic driven inlet valve and a
variable position outlet valve. A common variable position cooling tower bypass valve
exists in the system. The system circulates cooling tower bottom pond water through the
condenser and back to the top pond of the cooling tower. Currently the control system
runs with both pumps turned on DOL with the output valves open at 100 %. The cooling
tower bypass valve is used to control the condenser back pressure in the correct vacuum
range. The action of this valve allows some water to return directly to the bottom pond
and not be cooled by returning to the top pond. This in turn increases the temperature
of the bottom pond water that is to be circulated. The figure 6.2 shows the configuration
of the unit CW system at full load.
Below in table 6.11 is the current usage summary for the Circulating Water system running
at different generator load set points for 24 hours.
Below are the possible energy reduction strategies. Each option has three sections, the
strategy, the auxiliary MW result, and a risk comment.
6.4 Circulating Water Pumps 108
Figure 6.2: Circulating Water System Components - Tarong Power Station (Siemens 2015).
6.4 Circulating Water Pumps 109
Gen Load 150 175 200 225 250 275 300 325 350
CW Sys 64.2 65.4 66.3 66.8 67.6 68.8 69.2 69.1 69.6
Table 6.11: MATLAB Process Model Output - Circulating Water System Energy Usage
Figures at 9 Load Points to be used as a Baseline. Generator Load Set Point in MW and
Energy Totals in MWhrs.
Reduction Option 1
Strategy
Delay the start of the second pump until the unit as loaded up to the point of the process
requiring the two pumps. As the unit loads up use the condenser vacuum and bypass
valve position as an indicator to bring the second pump into service. On load down use
the bypass valve position as an indicator to remove the second pump from service.
Resultant Usage
Below in table 6.12 is the resultant usage summary for the Circulating Water system once
reduction option 1 is implemented using a conservative value of 200 MW to call the second
CW pump service requirement. This results in a reduction of 50 % energy consumption
in this plant area below 200 MW. The impact on the overall total auxiliary total is a 1
% reduction.
Gen Load 150 175 200 225 250 275 300 325 350
CW Sys 32.0 32.5 66.3 66.8 67.6 68.8 69.2 69.1 69.6
Table 6.12: MATLAB Process Model Output - Circulating Water System Energy Usage
Figures at 9 Load Points after Option 1 Implemented. Generator Load Set Point in MW and
Energy Totals in MWhrs
Risk Comment
The primary risk is that the second pump will fail to start when required holding up the
unit load up. This risk exists due to the CW hydraulic system requiring some upgrade
work. When this work is completed the call to start and stop a second CW pump would
be very low risk. Additional starts on the motor and pump set holds some increased risk
6.5 Condensate Extraction Pumps 110
compared to the current running arrangements.
Reduction Option 2
Strategy
Use the discharge valves to vary circulating water flow in line with the requirement of
condenser vacuum;
Resultant Usage
Below in table 6.13 is the resultant usage summary for the Circulating Water system
using a conservative value of 80 % for the reduction of loading below 200 MW. This was
difficult to accurately determine due to the lack of data available for these pumps. 80 %
was a figure determined based on generic pump load curves.
Gen Load 150 175 200 225 250 275 300 325 350
CW Sys 51.4 52.3 66.3 66.8 67.6 68.8 69.2 69.1 69.6
Table 6.13: MATLAB Process Model Output - Circulating Water System Energy Usage
Figures at 9 Load Points after Option 2 Implemented. Generator Load Set Point in MW and
Energy Totals in MW/hr
Risk Comment
This option is low risk as both pumps are in-service and the output valve can be manip-
ulated very quickly to give full flow to the condenser. Long term the increase wear is a
concern.
6.5 Condensate Extraction Pumps
This section is to look at potential ways in which possible energy reduction in the area of
CEPs could be achieved.
The system consists of two pumps that draw water from the condenser hotwell and dis-
charge it through the low pressure heaters into the deaerator. Each pump is driven by a
DOL motor with the outlet of pump controlled by a control valve. Currently the control
6.6 Boiler Feed Pumps 111
system runs with one pump in-service across the full load range.
Below in table 6.14 is the current usage summary for the Condensate Extraction system
running at different generator load set points for 24 hours.
Gen Load 150 175 200 225 250 275 300 325 350
CE Sys 16.0 16.7 17.4 18.2 18.8 19.9 20.7 20.7 21.5
Table 6.14: MATLAB Process Model Output - Condensate Extraction System Energy Usage
Figures at 9 Load Points to be used as a Baseline. Generator Load Set Point in MW and
Energy Totals in MWhrs.
There were no energy reduction strategies identified for this plant area.
6.6 Boiler Feed Pumps
This section is to look at potential ways in which possible energy reduction in the area of
BFPs could be achieved.
The system consists of three pumps that draw water from the deaerator and discharge it
through the high pressure heaters into the boiler drum. Each pump is driven by a DOL
motor with the outlet flow of pump controlled by a variable speed oil coupling. Currently
the control system runs with two pumps in-service across the full load range. The figure
6.3 shows the configuration of the unit BFP system at full unit load, 350 MW.
Below in table 6.15 is the current usage summary for the Boiler Feed Pump system running
at different generator load set points for 24 hours.
Gen Load 150 175 200 225 250 275 300 325 350
BFP Sys 93.1 103.8 121.0 131.3 150.1 171.4 189.2 198.5 217.6
Table 6.15: MATLAB Process Model Output - Boiler Feed Pump System Energy Usage
Figures at 9 Load Points to be used as a Baseline. Generator Load Set Point in MW and
Energy Totals in MWhrs.
Below are the possible energy reduction strategies. Each option has three sections, the
strategy, the auxiliary MW result, and a risk comment.
6.6 Boiler Feed Pumps 112
Figure 6.3: Boiler Feed Pump System Components - Tarong Power Station (Siemens 2015).
6.6 Boiler Feed Pumps 113
Reduction Option 1
Strategy
This reduction option is to run at low load below 200 MW with only one BFP in service
rather than two running inefficiently.
Resultant Usage
Below in table 6.16 is the resultant usage summary for the Boiler Feed Pump system
using only one BFP to supply the full feedwater flow up to 200 MW. This strategy results
in a 10 % reduction of energy consumption for this plant area under 200 MW. The closer
to 140 MW the more effective this strategy is with the reduction being 12 % at 150 MW
generator load set point.
Gen Load 150 175 200 225 250 275 300 325 350
BFP Sys 82.9 92.7 112.3 131.3 150.1 171.4 189.2 198.5 217.6
Table 6.16: MATLAB Process Model Output - Boiler Feed Pump System Energy Usage
Figures at 9 Load Points after Option 1 Implemented. Generator Load Set Point in MW and
Energy Totals in MWhrs.
Risk Comment
The primary risk is that the second pump will fail to start when required holding up the
unit load up. This risk can be considered very low as failure to start issue with a BFP
is a rare occurrence. The risk is mitigate further when the third pump is also available.
Additional starts on the motor and pump set holds some increased risk compared to the
current running arrangements.
Reduction Option 2
Strategy
This option is an extension to BFP reduction option 1. This reduction option is to run
at low load below 200 MW with only one BFP in service. Then at 200 MW the second
pump is started. At this point the first pump is to be locked in to the optimum efficiency
point with the second pump controlling the drum level.
6.6 Boiler Feed Pumps 114
Figure 6.4: Boiler Feed Pump Plant Area - Process Model versus Reduction Options
Resultant Usage
Below in table 6.17 is the resultant usage summary for the Boiler Feed Pump system using
only one BFP to supply the full feedwater flow up to 200 MW. Then an efficiency load
sharing option from this point upwards. This strategy results in a 0.46 % reduction of
energy consumption for this plant area across the full load range. The closer to 140 MW
the more effective this strategy is with the reduction being 12 % at 150 MW generator
load set point. A trend is included below in figure 6.4 to show the affect across the load
range.
Gen Load 150 175 200 225 250 275 300 325 350
BFP Sys 82.9 92.7 112.3 137.0 156.8 168.8 184.7 203.6 218.1
Table 6.17: MATLAB Process Model Output - Boiler Feed Pump System Energy Usage
Figures at 9 Load Points after Option 2 Implemented. Generator Load Set Point in MW and
Energy Totals in MWhrs.
Risk Comment
6.7 Reduction Options Summary 115
The primary risk is that the second pump will fail to start when required holding up the
unit load up. This risk can be considered very low as failure to start issue with a BFP
is a rare occurrence. The risk is mitigate further when the third pump is also available.
Additional starts on the motor and pump set holds some increased risk compared to
the current running arrangements. The pumps are designed to run continuously at full
efficiency and often do when the unit is fully loaded.
6.7 Reduction Options Summary
This section presents a summary of the reduction options tested on the process SIMULINK
model and presented in the sections above.
The impacts of each reduction option tested is presented in the table 6.18 below. This
table shows the impact of each proposed reduction options on the energy consumption of
the plant area involved. The resulting reduction is expressed as a percentage.
Gen Load MW 150 175 200 225 250 275 300 325 350
Option 1 CW % -50.2 -50.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Option 2 CW % -20.0 -20.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Option 1 BFP % -11.1 -10.8 -7.2 0.0 0.0 0.0 0.0 0.0 0.0
Option 2 BFP % -11.1 -10.8 -7.2 +4.3 +4.4 -1.5 -2.4 +2.6 +0.2
Option 1 ID % -10.0 -10.0 -10.0 -10.0 -10.0 -10.0 -10.0 -10.0 -10.0
Option 2 ID % -45.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Option 3 ID % -45.0 +0.2 -1.0 -1.1 -1.4 -1.2 -1.0 +1.2 0.0
Option 1 FD % -13.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Option 2 FD % -13.5 +0.2 +1.9 +3.7 +5.8 +9.6 +2.3 +3.7 0.0
Option 1 PA % -36.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Option 2 PA % -36.2 -9.5 +4.3 -0.7 +7.2 +18.9 +14.4 +22.4 +29.0
Table 6.18: Impact of Reduction Options in Each Plant Area at Different Generator Load Set
Points in MW - Resulting Impact Presented as a Percentage.
The impacts of each reduction option tested is presented in the table 6.18 below. This
table shows the impact of each proposed reduction options on the overall auxiliary energy
consumption. The resulting reduction is expressed as a percentage.
6.8 Entire Unit - Optimum Solution 116
Gen Load 150 175 200 225 250 275 300 325 350
Option 1 CW % -8.8 -8.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Option 2 CW % -3.5 -3.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Option 1 BFP % -2.8 -2.9 -2.1 0.0 0.0 0.0 0.0 0.0 0.0
Option 2 BFP % -2.8 0.0 0.0 +1.3 +1.4 -0.5 -0.9 +0.9 +0.1
Option 1 ID % -2.7 -2.6 -2.5 -2.5 -2.4 -2.2 -2.2 -2.1 -2.0
Option 2 ID % -11.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Option 3 ID % -11.4 +0.1 -0.3 -0.3 -0.3 -0.3 -0.2 +0.3 0.0
Option 1 FD % -1.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Option 2 FD % -1.3 +0.1 +0.2 +0.3 +0.6 +0.9 +0.2 +0.4 0.0
Option 1 PA % -2.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Option 2 PA % -2.5 -0.7 +0.3 -0.1 +0.5 +1.2 +0.9 +1.4 +1.7
Table 6.19: Impact of Reduction Options on the Overall Auxiliary Energy Consumption at
Different Generator Load Set Points in MW - Resulting Impact Presented as a Percentage.
6.8 Entire Unit - Optimum Solution
This section is to look at final combination of reduction options that gives the best final
resulting energy efficiency. The combinations presented below have taken into account
both the actual reduction achievable and the level of risk involved in implementation on
the plant.
The table 6.20 below contains a summary of the auxiliary power consumption totals at
different generator load set points for 24 hours. The data in this table represents the
baseline figure with which the proposed unit solutions can be compared.
Gen Load 150 175 200 225 250 275 300 325 350
Full Unit 368 392 417 435 465 497 528 548 578
kWhr/MW 102.09 93.28 86.77 80.58 77.54 75.24 73.27 70.25 68.86
Table 6.20: MATLAB Process Model Output - Full Unit Auxiliary Energy Usage Baseline
Figures at 9 Load Points. Generator Load Set Point in MW and Energy Totals in MWhrs.
6.8 Entire Unit - Optimum Solution 117
6.8.1 Low Risk
Optimal Unit Wide Reduction Option
Strategy
The low risk option to reduce the overall auxiliary power consumption is to implement
BFP reduction option 2 and CW reduction option 1 together. These options combine to
propose to run with only 1 BFP and 1 CW pump in service below 200 MW. Above 200
MW the circulating water system runs with the two pumps running and a second BFP
is called into service. The BFP’s are run with one at optimum loading and the second
modulating to maintain the required feedwater flow.
Resultant Usage
Below in table 6.21 is the resultant usage summary which presents the resulting full unit
24 hour auxiliary energy consumption after implementation of the low risk strategy. The
impact is compared to the current baseline running with the resulting change presented
as both a raw MWhr reduction and also as a percentage improvement. The final row
of the table presents the total kWhrs required to generate 1 MWhr of electricity at the
9 load set points. This low risk strategy gives an overall reduction in auxiliary energy
consumption of 10 % below 200 MW.
Gen Load 150 175 200 225 250 275 300 325 350
Full Unit 325 348 374 441 472 494 523 553 579
MWhr Impact -42.6 -44.1 -42.3 +5.6 +6.6 -2.6 -4.5 +5.1 +0.5
% Impact -11.6 -11.2 -10.2 +1.3 +1.4 -0.5 -0.9 +0.9 +0.1
kWhr/1MWhr 90.27 82.79 77.96 81.63 78.64 74.85 72.64 70.91 68.92
Table 6.21: MATLAB Process Model Output - Full Unit Energy Usage Figures at 9 Load
Points after Low Risk Options are Implemented. Generator Load Set Point in MW and
Energy Totals in MWhrs.
6.8.2 Medium Risk
Optimal Unit Wide Reduction Option
6.8 Entire Unit - Optimum Solution 118
Strategy
The medium risk strategy is an extension of the low risk strategy above. This strategy
is to implement the low risk drive running modifications and also to increase the furnace
pressure set point closer to the zero to reduce the load on the ID fans, reduction option
1. This set point change will reduce the ID fan load by 10% across the full load range of
the unit.
Resultant Usage
Below in table 6.22 is the resultant usage summary which presents the resulting full unit 24
hour auxiliary energy consumption after implementation of the medium risk strategy. The
impact is compared to the current baseline running with the resulting change presented
as both a raw MWhr reduction and also as a percentage improvement. The final row of
the table presents the total kWhrs required to generate 1 MWhr of electricity at the 9
load set points. This medium risk strategy gives an overall reduction in auxiliary energy
consumption of 10 - 15 % below 200 MW and a modest reduction of 1 - 3 % across the
remaining load range.
Gen Load 150 175 200 225 250 275 300 325 350
Full Unit 315 338 364 430 461 483 512 542 567
MWhr Impact -52.3 -54.2 -52.8 -5.0 -4.3 -13.7 -16.0 -6.4 -11.2
% Impact -14.2 -13.8 -12.7 -1.2 -0.9 -2.8 -3.0 -1.2 -1.9
kWhr/1MWhr 87.57 80.38 75.77 79.65 76.81 73.17 71.05 69.43 67.52
Table 6.22: MATLAB Process Model Output - Full Unit Energy Usage Figures at 9 Load
Points after Low and Medium Risk Options are Implemented. Generator Load Set Point in
MW and Energy Totals in MWhrs.
6.8.3 High Risk
Optimal Unit Wide Reduction Option
Strategy
The high risk strategy is an extension of the medium risk strategy above. This strategy
6.8 Entire Unit - Optimum Solution 119
is to implement the low risk drive running modifications, increase the furnace pressure
set point and also run with only one half of the fan groups in service until the process
requires them. This is reduction options ID fan 3, FD fan 1 and PA fan 1. This means
that until 175 MW only 1 CW pump, BFP, ID fan, FD fan and PA fan are in service.
At 175MW the second ID, FD and PA fans are started. The second CW pump and BFP
are brought into service at 200 MW. When the second fan groups starts ID fan loading
will be manipulated to share the load to ensure one drive is always optimally loaded.
Once the second FD and PA fans are started the loading will be shared as the modeling
investigation has found that to be the ideal energy efficiency strategy. To reduce the long
term chance of uneven wear rates within the air and gas paths it would be proposed to
alternate the fan group that is removed from service at low load. A high level monitoring
sequence to ensure long term run hours are matched would also be required.
Resultant Usage
Below in table 6.23 is the resultant usage summary which presents the resulting full unit
24 hour auxiliary energy consumption after implementation of the high risk strategy. The
impact is compared to the current baseline running with the resulting change presented
as both a raw MWhr reduction and also as a percentage improvement. The final row
of the table presents the total kWhrs required to generate 1 MWhr of electricity at the
9 load set points. This high risk strategy gives an overall reduction in auxiliary energy
consumption of 15 - 10 % below 200 MW and a reduction of 3 - 5 % across the remaining
load range.
Gen Load 150 175 200 225 250 275 300 325 350
Full Unit 259 335 355 421 451 473 501 527 551
MWhr Impact -108.5 -56.8 -61.9 -14.5 -14.3 -23.8 -26.2 -20.6 -27.3
% Impact -29.5 -14.5 -14.9 -3.3 -3.1 -4.8 -5.0 -3.8 -4.7
kWhr/1MWhr 71.96 79.76 73.89 77.89 75.15 71.64 69.62 67.61 65.61
Table 6.23: MATLAB Process Model Output - Full Unit Energy Usage Figures at 9 Load
Points after Low, Medium and High Risk Options are Implemented. Generator Load Set
Point in MW and Energy Totals in MWhrs.
6.8 Entire Unit - Optimum Solution 120
Figure 6.5: Full Unit Auxiliary Energy Consumption Comparison No 1 - Baseline versus Risk
Reduction Options.
6.8.4 Reduction Summary
The two trends included in figure 6.5 and 6.6 to show the affect across the load range of
the 3 unit wide solutions against the current running baseline.
6.8 Entire Unit - Optimum Solution 121
Figure 6.6: Full Unit Auxiliary Energy Consumption Comparison No 2 - Baseline versus Risk
Reduction Options.
6.9 Conclusion 122
6.9 Conclusion
In this chapter the potential options for energy reduction through operational methods
were explored. Each plant area was explored individually with an overall unit wide ideal
solution proposed. Where possible the options proposed were theoretically implemented
utilising the created MATLAB models to measure the energy impact of the option. The
risk discussions have given grounding for broader discussions that may lead to full im-
plementation on running plant. The next chapter outlines the finial conclusions of the
project and the potential future work.
Chapter 7
Conclusion and Further Work
This chapter is to present the projects achievements, key findings and to outline any
opportunities for future work.
The project has successfully achieved the projects two primary aims of understanding
the auxiliary energy consumption at TPS and to identify potential control based energy
reduction strategies. The understanding of the auxiliary usage was achieved through
the completion of a comprehensive literature review, the successful creation of energy
tracking logic within the plant control system, Siemens T3000, and through the creation
of multiple MATLAB models. The process of identification of zero capital reduction
strategies utilised the knowledge gained during the completion of the primary aims and
the knowledge of the author. A memo to Stanwell regarding potential energy reduction
strategies and the steps required to implement was created. The creation of T3000 logic
achieved the projects secondary aim of leaving in place energy consumption information
that could be utilised into the future. The T3000 logic created will allow ongoing realtime
monitoring of auxiliary consumption within TPS.
7.1 Project Conclusions
The project successfully created multiple auxiliary energy models in MATLAB. The mod-
els were based on data that was extracted from the TPS data historian, PIMS. These
models were utilised to understand the auxiliary energy usage at TPS. The understand-
7.1 Project Conclusions 124
ing and model accuracy was proven against the T3000 energy tracking logic that the
project also created. The T3000 logic was in turn confirmed against the permanently
installed energy meters. This proved that an accurate model of coal fired power station
auxiliary energy consumption is able to be created based on data mining techniques.
The project has concluded that reductions of auxiliary energy consumption at Tarong
Power Station is achievable without capital investment. This conclusion is supported by
the understanding gained through the comprehensive literature review completed, the
creation of T3000 tracking logic, and working MATLAB models. The ability to achieve
reduction was confirmed through the use of the created MATLAB SIMULINK process
model to successfully load share and prove the auxiliary power reduction. The proposed
unit wide reduction options, including results of the SIMULINK trials, is discussed in
chapter 6.
A summary and steps to implement the recommended reduction strategies on the actual
TPS unit plant are included in Appendix I Memo to Engineering Superintendent - Elec-
trical Maintenance - TPS. This memo also outlines that the reduction options proposed
can be applied to all 4 Tarong Power Station units and that there is a potential that
the same strategies could be applied to the 4 Stanwell Power Station units for similar
gains. This is due to the fact that Stanwell Power Station design and load profiles are
very similar to Tarong. This can be extended to say that the project has concluded that
any unit with a similar design could benefit from operational improvement based energy
efficiency strategies, particulary at low loads. All three possible options are outline below
with the Low and Medium risk options being proposed to be implemented within TPS
unit plant.
The final proposed low risk unit wide optimal reduction strategy has 2 key components;
1. Circulating Water pump control modification - only one CW pump in-service until
the process requirements call for the second pump - approximately 200 MW; and
2. Boiler Feed Pump control modification - only 1 BFP in-service until the process
calls for it - approximately 200 MW - a optimum trim solution up to 300 MW will
also give some benefit.
This low risk strategy gives an overall reduction in auxiliary energy consumption of 10
7.1 Project Conclusions 125
% below 200 MW. The overall reduction will vary with load profile, ambient weather
conditions and status of the plant in the maintenance cycle. The CW and BFP drive
control modifications only give benefit in the low load range. The implementation of
these reduction strategies is considered low risk as the primary risk is that the second
drive will not start in the time required meaning that the unit load up may be delayed.
The project also found that if a medium risk profile was taken on then additional reduction
could be achieved. The additional item below needs to be added to the low risk strategy
above;
1. Induced Draft Fan control modification - lifting the furnace pressure set point closer
to zero.
The ID fan control modification will be effective across the full load range with an in-
creased risk that the boiler may trip on furnace pressure excursion. This risk is of particu-
lar concern when a clinker fall occurs and the pressure ripple it creates forces the pressure
high for greater than the 2 second delay time. This is the reason for the medium risk
rating of implementing this energy reduction strategy. This medium risk strategy gives
an overall reduction in auxiliary energy consumption of 10 - 15 % below 200 MW and a
modest reduction of 1 - 3 % across the remaining load range.
The project also found that if a higher risk profile was taken on then additional reduction
could be achieved. The additional items below need to be added to the medium risk
strategy above;
1. Primary Air fan control modification - only one PA fan in-service until the process
requirements call for the second fan - approximately 175 MW;
2. Forced Draft fan control modification - only one FD fan in-service until the process
requirements call for the second fan - approximately 175 MW; and
3. Induced Draft fan control modification - only one ID fan in-service until the process
requirements call for the second fan - approximately 175 MW. Once the second fan
is in service some load manipulation gives additional consumption reduction across
the remaining load range.
7.2 Future Work 126
These 3 higher risk control modifications would also require a duty/standby fan arrange-
ment to be implemented to ensure that wear rates are balanced out. It would be proposed
that only half the air and gas group be in-service up until 175 MW. Each time the load
cycles into the low load range the opposite side to last is removed from service. This would
assist in wear rate balance but a higher level runtime monitoring system would need to
sit over the top of this basic duty/standby control proposed to ensure long term balance.
This high risk strategy gives an overall reduction in auxiliary energy consumption of 15
- 10 % below 200 MW and a reduction of 3 - 5 % across the remaining load range.
The project has also found another key relating to auxiliary energy consumption at TPS
is that over 300 MW the unit is running at peak auxiliary power efficiency and there is
little that can be done to reduce the energy consumption rate. Attempts to share drive
loading at the higher generator set points have little to no positive effect. As discussed
earlier the ID fan load can be reduced across the full load range by lifting the furnace set
point.
While this project has made significant contribution to the industry there are items iden-
tified during this project that require additional work to complete, these will be outlined
in the next section.
7.2 Future Work
This section is to outline potential future work that may use this projects findings as a
basis.
The project has identified future work that would further benefit the Tarong Power Station
and the broader electricity industry. This future work is broadly grouped into 3 categories,
MATLAB, T3000 and Other. These potential areas for future work are outlined below.
MATLAB
The items below are possible enhancements that could be made to the created MATLAB
models to assist in continued expansion of the knowledge of auxiliary energy consumption.
The models could be expanded or modified to allow other plants to be more easily assessed.
7.2 Future Work 127
• addition of drive service select inputs similar to the BFP selections;
• creation of live links to the plant data from the MATLAB model to allow manipu-
lation of the load distribution in realtime;
• modify the models to make it more generic, so they can potentially be applied more
easily to other plants;
• addition of 415 V drives to the model to allow more detailed investigation of the im-
pacts of drives at that level. Examples of these drives are mill motors, air extraction
pumps and flame detection cooling fans;
• creation of not in-service switches for all drives similar to the BFP drives.
T3000
The items below will expand on the current work in T3000 by making the created data
more visible and potentially implementing the recommended energy reduction strategies.
• creation of auxiliary energy flow HMI;
• creation of auxiliary energy consumption alarms, realtime and 24 hour moving av-
erage;
• creation of auxiliary energy trends;
• BFP transfer logic at low loads;
• CW transfer logic at low loads;
• creation of boiler furnace ride though logic to allow the furnace set point to be lifted;
• BFP trim logic to ensure one pump is always at optimal.
Other
The items below will investigate those items identified during the literature review that
were outside of the scope of this project. These items are viewed as having potential
industry benefit but were either to wide ranging or requiring capital injection to be con-
sidered under this project. All of these items could be investigated from the perspective
of auxiliary energy reduction and life cycle of the components.
7.3 Conclusion 128
• review the findings at Tarong for potential implementation at Stanwell Power Sta-
tion;
• investigate ID fan impeller hard facing to reduce wear rate and overall loading;
• investigate ASD’s for the BFP’s to reduce plant area energy consumption;
• investigate hard facing of milling components, balls and rings to reduce wear rate
and overall loading;
• investigate the potential of reducing size of coal entering the pulveriser to reduce
overall loading;
• investigate the energy savings in routinely sweeping the mills to reduce mill loading;
• investigate the impact of air heater overhaul intervals on ID fan loading;
• investigation of ammonia injection into the flu gas to reduce precipitator pressure
drop and in turn ID fan loading;
• investigation of installation of power factor correction devices.
Many suggestions above would make ideal stand alone projects or as a combined project
to look at reduction of auxiliary energy consumption through capital investment which is
the opposite perspective of this projects aims.
7.3 Conclusion
In this chapter the projects achievements, conclusions and opportunities for future work
were outlined. The project has concluded that reductions of auxiliary energy consumption
at Tarong Power Station is achievable without capital investment. This energy efficiency
improvement can be achieved solely through control optimisation using the existing in-
stalled infrastructure.
Bibliography
Alstrom (1981), ‘A1-V-503115-01 Unit 6.6kV Board Switchgear Detail Diagram Tarong
Power Station’, QEGB Switchboard Manual .
Asumadu, J., Devaney, M., Wallis, L. & Bond, J. (1990), ‘Industrial Power Measurement
and Control Design’, Unknown .
Australia, E. (2015), ‘Chronologies of Engineering Works in Queensland’, https:
//www.engineersaustralia.org.au/engineering-heritage-queensland/
chronology-engineering-works-queensland. [Online; accessed May-2015].
Boveri, A. B. (2015), ‘Making Power Plants Energy Efficient’, http://abb. [Online;
accessed May-2015].
Boylestad, R. L. (2003), Introductory Circuit Anlysis, Prentice Hall. Tenth Edition.
Brady, F. (1996), ‘A Dictionary on Electricity - A Contribution on Australia’,
http://www.ewh.ieee.org/r10/nsw/subpages/history/electricity_in_
australia.pdf. [Online; accessed May-2015].
Breda, R. V. (2014), ‘Stanwell Annual Report 2014’, http://www.stanwell.com/story/
annual-reports/. [Online; accessed May-2015].
Changliang, L., Hong, W., Jinliang, D. & Chenggang, Z. (2011), ‘An Overview of Mod-
elling and Simulation of Thermal Power Plant’, The 2011 International Conference
on Advanced Mechatronic Systems .
Changliang, L., Jizhen, L., Yuguang, N. & Xiuzhang, J. (2004), ‘Nonlinear Modeling and
Simulation for Large Scale Coal-Fired Power Unit’, The 30th Annual Conference of
the IEEE Industrial Society .
BIBLIOGRAPHY 130
Cornerstone (2014), ‘Coal-Based Electricity Generation in India’, http:
//cornerstonemag.net/coal-based-electricity-generation-in-india/.
[Online; accessed May-2015].
Economy, R. (2013), ‘Stanwell Blames Solar for Decline in
Fossil Fuel Baseload’, http://reneweconomy.com.au/2013/
stanwell-blames-solar-for-decline-in-fossil-fuel-baseload-54543.
[Online; accessed May-2015].
EPRI, E. P. R. I. (2011), ‘Oppurtunities to Enhance Electric Energy Efficiency in the Pro-
duction and Delivery of Electricity’, Program on Technology Innovation: Electricity
Use in the Electric Sector . Technical Report.
Godsmark, M. (2001), Energy Metering Installation Manual, Energy Metering Manual,
Electrix.
Government, Q. (2011), Work Health and Safety Act 2011, Standard WHS ACT:2011,
Division of Workplace Health Safety.
Hitachi (2003), Operations and Maintenance Manuals, Boiler and Turbine Manuals, Hi-
tachi.
IEEE (2013), Recommended Practice for the Instrumentation and Metering of Indus-
trial and Commercial Power Systems, Standard IEEE Std 3001.8:2013, Institute of
Electrical and Electronics Engineers.
Lang, J. (2015), ‘Personal Experience’, Informally. [20 Years Power Generation Industry
Experience].
Li, Y., Wang, J.-J., Jiang, T.-L. & Zhang, B.-W. (2009), ‘Energy Audit and Its Applica-
tion in Coal-fired Power Plant’, IEEE Science and Technology Department Plans of
Jilin Province of China .
Ltd, S. P. G. P. (2015), ‘T3000 Product Overview’, http://www.energy.siemens.com/
hq/en/automation/power-generation/sppa-t3000.htm. [Online; accessed May-
2015].
Mandi, P., Rajasheker, M., Udaykumar, D. & Yaragatti, R. (2012), ‘Energy Efficiency
Improvement of Auxiliary Power Equipment in Thermal Power Plant through Oper-
ational Optimization’, IEEE International Conference on Power Electronics, Drives
and Energy Systems .
BIBLIOGRAPHY 131
Ningling, W., Yongping, Y. & Zhiping, Y. (2009), ‘Energy-saving Analysis for a 600mw
Coal-Fired Supercritical Power Plant’, National Basic Research Program of China .
Ningling, W., Yongping, Y., Zhiping, Y. & Yu, W. (2010), ‘Diagnosis of Energy-saving
Potential and Optimized Measures for 600mw Supercritical Coal-Fired Power Unit’,
2010 International Conference of Electrical and Control Engineering .
Sands, M. & Blake, P. (2015), ‘Tarong Power Station Monthly Performance Report’,
Internal Monthly Report .
Schwanz, D., Silva, M., Leborgne, R., Bretas, A. & Gaidzinski, M. (2014), ‘Improve-
ment of a Coal Power Plant Performance Through Auxiliary Service Power Demand
Reduction’, IEEE .
Siemens (2015), ‘T3000 Unit Overview’, Tarong Power Station DCS. [Live Plant Data;
accessed 2015].
Simon, E., Stoica, C., Rodriguez-Ayerbe, P., Dumur, D. & Wertz, V. (2010), ‘Robustified
Optimal Control of a Coal-fired Power Plant’, American Control Conference .
Thiess (2015), ‘Tarong Power Station Siteworks’, https://www.thiess.com/projects/
tarong-power-station-siteworks. [Online; accessed May-2015].
Yang, Y.-P., Wang, N.-L., Zhang, Z.-W. & Chen, D.-G. (2010), ‘Data Mining-Based
Modeling and Application in the Energy-Saving Analysis of Large Coal-Fired Power
Units’, Ninth International Conference of Machine Learning and Cybernetics .
Zecevic, M. & EEUAA (2015), ‘Australian National Electricity Market Report - Week 1
August 2015’, Energy User Association of Australia Weekly Report .
Appendix A
Project Specification
ENG 4111/2 Research Project
Project Specification
For: Jason Lang
Topic: Energy Efficiency Improvement in Coal Fired Power Plant through
Operational Optimisation
Supervisors: Catherine Hills
Sponsorship: Stanwell Corporation Limited
Project Aim: This project aims to model the auxiliary energy usage of a 350MW
coal fired unit using MATLAB and T3000 Digital Control System
logic. Using this model it aims to propose control methods to re-
duce the internal electrical energy consumed to produce electricity.
Program:
1. Research into auxiliary energy, energy calculations, energy reduction and control
optimisation in coal fired power stations.
2. Collect energy data from existing plant operations and maintenance manuals.
3. Collect historical plant energy usage data for all unit conditions.
4. Create a SIMULINK model, in MATLAB, of the energy profile of a single unit.
5. Create energy tracking logic in the T3000 Digital Control System.
133
6. Use local energy meter readings on Unit 4 to confirm/tune Digital Control System
tracking values.
7. Calibrate the results of the MATLAB model against the real overall measured power
for the unit.
8. Use the model to determine optimal solutions for energy reduction through control
optimisation.
9. Select the solutions with acceptable risk and good potential and analyse these
schemes for practicality and viability.
10. Propose control system modifications and operational guidelines in a report to be
delivered to site.
As time, resources and sponsor plant risk assessment permit:
1. Trial optimised energy scheme on the unit simulator and evaluate.
2. Implement optimised energy scheme on a running unit and evaluate.
Appendix B
Conceptual MATLAB Auxiliary
Calculation Script
135
%This file is to test the power calculation result for 1 full day
%All voltage and current measurements have been exported from PIMS and
%then imported into MATLAB.
%Once imported the data is split into data arrays and the real power is
%then calculated.
%The results are roughly displayed on the workspace.
%
%Clean up workspace
clear all
close all
%
%Setup blank matrices
row = 1;
%
%Import data from excel spreadsheet
[numdata, txtdata, rawdata] = xlsread('U47thJulyAux.xlsx');
%
%Manipulate column numerical data into individual vectors
GEN_20KV_V = numdata(1:end,1); %Generator 20kV Bus Volts
MW = numdata(1:end,2); %MW's Generated
A_UNIT_TX_TOT_MW = numdata(1:end,3); %'A' Unit Transformer MW
A_UNIT_TX_TOT_PF = numdata(1:end,4); %'A' Unit Transformer Power Factor
A_UNIT_TX_AVG_I = numdata(1:end,5); %'A' Unit Transformer Average Current
B_UNIT_TX_TOT_MW = numdata(1:end,6); %'B' Unit Transformer MW
B_UNIT_TX_TOT_PF = numdata(1:end,7); %'B' Unit Transformer Power Factor
B_UNIT_TX_AVG_I = numdata(1:end,8); %'B' Unit Transformer Average Current
A_6_6_BUS_V = numdata(1:end,9); %'A' 6.6kV Bus Volts
A_6_6_BUS_I = numdata(1:end,10); %'A' 6.6kV Bus Current
B_6_6_BUS_V = numdata(1:end,11); %'B' 6.6kV Bus Volts
B_6_6_BUS_I = numdata(1:end,12); %'B' 6.6kV Bus Current
C_6_6_BUS_V = numdata(1:end,13); %'C' 6.6kV Bus Volts
BNKR_TX_6_6_I = numdata(1:end,14); %Bunkering Transformer Current
A_PRECIP_TX_6_6_I = numdata(1:end,15);%'A' Precip Transformer Current
B_PRECIP_TX_6_6_I = numdata(1:end,16);%'B' Precip Transformer Current
A_415V_UNIT_TX_I = numdata(1:end,17); %'A' 415V Unit Transformer Current
B_415V_UNIT_TX_I = numdata(1:end,18); %'B' 415V Unit Transformer Current
C_415V_UNIT_TX_I = numdata(1:end,19); %'C' 415V Unit Transformer Current
A_PA_FAN_LD = numdata(1:end,20); %'A' Primary Air Fan Motor Load
B_PA_FAN_LD = numdata(1:end,21); %'B' Primary Air Fan Motor Load
A_FD_FAN_LD = numdata(1:end,22); %'A' Forced Draft Air Fan Motor Load
B_FD_FAN_LD = numdata(1:end,23); %'B' Forced Draft Air Fan Motor Load
A_ID_FAN_LD = numdata(1:end,24); %'A' Induced Draft Air Fan Motor Load
B_ID_FAN_LD = numdata(1:end,25); %'B' Induced Draft Air Fan Motor Load
A_BFP_LD = numdata(1:end,26); %'A' Boiler Feed Pump Motor Load
B_BFP_LD = numdata(1:end,27); %'B' Boiler Feed Pump Motor Load
C_BFP_LD = numdata(1:end,28); %'C' Boiler Feed Pump Motor Load
A_CEP_LD = numdata(1:end,29); %'A' Condensate Extraction Pump Motor Load
B_CEP_LD = numdata(1:end,30); %'B' Condensate Extraction Pump Motor Load
A_CW_LD = numdata(1:end,31); %'A' Circulating Water Pump Motor Load
B_CW_LD = numdata(1:end,32); %'B' Circulating Water Pump Motor Load
%
%Calulate fixed variables
sqrt3 = sqrt(3);
LV_V = 415;
%
%E(kWh) = P(W) x t(hr)/1000
136
%
%Calculate the daily power for 'A' Unit Transformer
A_PWR = sqrt3.*(GEN_20KV_V*1000).*A_UNIT_TX_AVG_I.*A_UNIT_TX_TOT_PF;
A_PWR_TOTAL = (sum(A_PWR)./1000000)/3600 %Total Calc 'A' Unit Transformer
A_PWR_MET_TOTAL = sum(A_UNIT_TX_TOT_MW) %Total Meas 'A' Unit Transformer
A_PWR_ERROR = ((A_PWR_TOTAL - A_PWR_MET_TOTAL)/A_PWR_MET_TOTAL)*100
%
%Calculate the daily power for 'B' Unit Transformer
B_PWR = sqrt3.*(GEN_20KV_V*1000).*B_UNIT_TX_AVG_I.*B_UNIT_TX_TOT_PF;
B_PWR_TOTAL = (sum(B_PWR)./1000000)/3600 %Total Calc 'B' Unit Transformer
B_PWR_MET_TOTAL = sum(B_UNIT_TX_TOT_MW) %Total Meas 'B' Unit Transformer
B_PWR_ERROR = ((B_PWR_TOTAL - B_PWR_MET_TOTAL)/B_PWR_MET_TOTAL)*100
%
%Calculate the daily power for Bunkering Transformer
BUN_PWR = sqrt3.*LV_V.*BNKR_TX_6_6_I.*A_UNIT_TX_TOT_PF;
BUN_PWR_TOTAL = (sum(BUN_PWR)./1000000)/3600 %Total Calc Bunkering Transf
%
%Calculate the daily power for 'A' Precip Transformer
A_PRE_PWR = sqrt3.*(A_6_6_BUS_V*1000).*A_PRECIP_TX_6_6_I.*A_UNIT_TX_TOT_PF;
A_PRE_PWR_TOTAL = (sum(A_PRE_PWR)./1000000)/3600%Total Calc A Precip Transf
%
%Calculate the daily power for 'B' Precip Transformer
B_PRE_PWR = sqrt3.*(B_6_6_BUS_V*1000).*B_PRECIP_TX_6_6_I.*B_UNIT_TX_TOT_PF;
B_PRE_PWR_TOTAL = (sum(B_PRE_PWR)./1000000)/3600%Total Calc B Precip Transf
%
%Calculate the daily power for 'A' 415V Unit Transformer
A_415V_PWR = sqrt3.*LV_V.*A_415V_UNIT_TX_I.*A_UNIT_TX_TOT_PF;
A_415V_PWR_TOTAL = (sum(A_415V_PWR)./1000000)/3600%Total Calc A 415V Transf
%
%Calculate the daily power for 'B' 415V Unit Transformer
B_415V_PWR = sqrt3.*LV_V.*B_415V_UNIT_TX_I.*B_UNIT_TX_TOT_PF;
B_415V_PWR_TOTAL = (sum(B_415V_PWR)./1000000)/3600%Total Calc B 415V Transf
%
%Calculate the daily power for 'C' 415V Unit Transformer
C_415V_PWR = sqrt3.*LV_V.*C_415V_UNIT_TX_I.*A_UNIT_TX_TOT_PF;
C_415V_PWR_TOTAL = (sum(C_415V_PWR)./1000000)/3600%Total Calc C 415V Transf
%
%Calculate the daily power for 'A' Primary Air Fan
A_PA_PWR = sqrt3.*(A_6_6_BUS_V*1000).*A_PA_FAN_LD.*A_UNIT_TX_TOT_PF;
A_PA_PWR_TOTAL = (sum(A_PA_PWR)./1000000)/3600 %Total Calc A PA Fan Load
%
%Calculate the daily power for 'B' Primary Air Fan
B_PA_PWR = sqrt3.*(B_6_6_BUS_V*1000).*B_PA_FAN_LD.*A_UNIT_TX_TOT_PF;
B_PA_PWR_TOTAL = (sum(B_PA_PWR)./1000000)/3600 %Total Calc B PA Fan Load
%
%Calculate the daily power for 'A' Forced Draft Fan
A_FD_PWR = sqrt3.*(A_6_6_BUS_V*1000).*A_FD_FAN_LD.*A_UNIT_TX_TOT_PF;
A_FD_PWR_TOTAL = (sum(A_FD_PWR)./1000000)/3600 %Total Calc A FD Fan Load
%
%Calculate the daily power for 'B' Forced Draft Fan
B_FD_PWR = sqrt3.*(B_6_6_BUS_V*1000).*B_FD_FAN_LD.*B_UNIT_TX_TOT_PF;
B_FD_PWR_TOTAL = (sum(B_FD_PWR)./1000000)/3600 %Total Calc B FD Fan Load
%
%Calculate the daily power for 'A' Induced Draft Fan
A_ID_PWR = sqrt3.*(A_6_6_BUS_V*1000).*A_ID_FAN_LD.*A_UNIT_TX_TOT_PF;
A_ID_PWR_TOTAL = (sum(A_ID_PWR)./1000000)/3600 %Total Calc A ID Fan Load
137
%Calculate the daily power for 'B' Induced Draft Fan
B_ID_PWR = sqrt3.*(B_6_6_BUS_V*1000).*B_ID_FAN_LD.*B_UNIT_TX_TOT_PF;
B_ID_PWR_TOTAL = (sum(B_ID_PWR)./1000000)/3600 %Total Calc B ID Fan Load
%
%Calculate the daily power for 'A' Boiler Feed Pump
A_BFP_PWR = sqrt3.*(A_6_6_BUS_V*1000).*A_BFP_LD.*A_UNIT_TX_TOT_PF;
A_BFP_PWR_TOTAL = (sum(A_BFP_PWR)./1000000)/3600 %Total Calc A BFP Load
%
%Calculate the daily power for 'B' Boiler Feed Pump
B_BFP_PWR = sqrt3.*(B_6_6_BUS_V*1000).*B_BFP_LD.*B_UNIT_TX_TOT_PF;
B_BFP_PWR_TOTAL = (sum(B_BFP_PWR)./1000000)/3600 %Total Calc B BFP Load
%
%Calculate the daily power for 'C' Boiler Feed Pump
C_BFP_PWR = sqrt3.*(A_6_6_BUS_V*1000).*C_BFP_LD.*A_UNIT_TX_TOT_PF;
C_BFP_PWR_TOTAL = (sum(C_BFP_PWR)./1000000)/3600 %Total Calc C BFP Load
%
%Calculate the daily power for 'A' Boiler Feed Pump
A_CEP_PWR = sqrt3.*(A_6_6_BUS_V*1000).*A_CEP_LD.*A_UNIT_TX_TOT_PF;
A_CEP_PWR_TOTAL = (sum(A_CEP_PWR)./1000000)/3600 %Total Calc A CEP Load
%
%Calculate the daily power for 'B' Boiler Feed Pump
B_CEP_PWR = sqrt3.*(B_6_6_BUS_V*1000).*B_CEP_LD.*B_UNIT_TX_TOT_PF;
B_CEP_PWR_TOTAL = (sum(B_CEP_PWR)./1000000)/3600 %Total Calc B CEP Load
%
%Calculate the daily power for 'A' Circulating Water Pump
A_CW_PWR = sqrt3.*(A_6_6_BUS_V*1000).*A_CW_LD.*A_UNIT_TX_TOT_PF;
A_CW_PWR_TOTAL = (sum(A_CW_PWR)./1000000)/3600 %Total Calc A CW Load
%
%Calculate the daily power for 'B' Circulating Water Pump
B_CW_PWR = sqrt3.*(B_6_6_BUS_V*1000).*B_CW_LD.*B_UNIT_TX_TOT_PF;
B_CW_PWR_TOTAL = (sum(B_CW_PWR)./1000000)/3600 %Total Calc B CW Load
%
%Calculate the 'A' bus load totals
A_BUS_LD_TOTAL = BUN_PWR_TOTAL + A_PRE_PWR_TOTAL + A_415V_PWR_TOTAL...
+ C_415V_PWR_TOTAL + A_PA_PWR_TOTAL + A_FD_PWR_TOTAL + A_ID_PWR_TOTAL...
+ A_BFP_PWR_TOTAL + C_BFP_PWR_TOTAL + A_CEP_PWR_TOTAL + A_CW_PWR_TOTAL
A_PWR_ERROR_2 = ((A_PWR_TOTAL - A_BUS_LD_TOTAL)/A_BUS_LD_TOTAL)*100
%
%Calculate the 'B' bus load totals
B_BUS_LD_TOTAL = B_PRE_PWR_TOTAL+ B_415V_PWR_TOTAL + B_PA_PWR_TOTAL...
+ B_FD_PWR_TOTAL + B_ID_PWR_TOTAL + B_BFP_PWR_TOTAL + B_CEP_PWR_TOTAL...
+ B_CW_PWR_TOTAL
B_PWR_ERROR_2 = ((B_PWR_TOTAL - B_BUS_LD_TOTAL)/B_BUS_LD_TOTAL)*100
%
MW_GEN = sum(MW)/3600
AUX_TOTAL = A_PWR_TOTAL + B_PWR_TOTAL
A_PWR_TOTAL =
240.3747
A_PWR_MET_TOTAL =
138
8.6079e+05
A_PWR_ERROR =
-99.9721
B_PWR_TOTAL =
226.3892
B_PWR_MET_TOTAL =
8.1075e+05
B_PWR_ERROR =
-99.9721
BUN_PWR_TOTAL =
0.1228
A_PRE_PWR_TOTAL =
0.8783
B_PRE_PWR_TOTAL =
1.1044
A_415V_PWR_TOTAL =
27.9323
B_415V_PWR_TOTAL =
7.0143
C_415V_PWR_TOTAL =
8.6071
A_PA_PWR_TOTAL =
14.6317
139
B_PA_PWR_TOTAL =
14.4754
A_FD_PWR_TOTAL =
21.7526
B_FD_PWR_TOTAL =
21.5251
A_ID_PWR_TOTAL =
54.4733
B_ID_PWR_TOTAL =
54.1522
A_BFP_PWR_TOTAL =
0.0645
B_BFP_PWR_TOTAL =
73.5466
C_BFP_PWR_TOTAL =
72.2956
A_CEP_PWR_TOTAL =
0
B_CEP_PWR_TOTAL =
18.7002
A_CW_PWR_TOTAL =
33.2420
B_CW_PWR_TOTAL =
34.1969
140
A_BUS_LD_TOTAL =
234.0003
A_PWR_ERROR_2 =
2.7241
B_BUS_LD_TOTAL =
224.7152
B_PWR_ERROR_2 =
0.7450
MW_GEN =
5.8192e+03
AUX_TOTAL =
466.7639
Appendix C
Unit Auxiliary Power
Consumption Simulink Model -
Power Model
142
Figure C.1: Unit Auxiliary Power Consumption Simulink Model - Power Model Part 1
143
Figure C.2: Unit Auxiliary Power Consumption Simulink Model - Power Model Part 2
144
Figure C.3: Unit Auxiliary Power Consumption Simulink Model - Power Model Part 3
145
Figure C.4: Unit Auxiliary Power Consumption Simulink Model - Power Model Part 4
146
Figure C.5: Unit Auxiliary Power Consumption Simulink Model - Power Model Part 5
147
Figure C.6: Unit Auxiliary Power Consumption Simulink Model - Power Model Part 6
148
Figure C.7: Unit Auxiliary Power Consumption Simulink Model - Power Model Part 7
Appendix D
Unit Auxiliary Power
Consumption Simulink Model -
Position Model
150
Figure D.1: Unit Auxiliary Power Consumption Simulink Model - Position Model Part 1
151
Figure D.2: Unit Auxiliary Power Consumption Simulink Model - Position Model Part 2
152
Figure D.3: Unit Auxiliary Power Consumption Simulink Model - Position Model Part 3
153
Figure D.4: Unit Auxiliary Power Consumption Simulink Model - Position Model Part 4
154
Figure D.5: Unit Auxiliary Power Consumption Simulink Model - Position Model Part 5
155
Figure D.6: Unit Auxiliary Power Consumption Simulink Model - Position Model Part 6
156
Figure D.7: Unit Auxiliary Power Consumption Simulink Model - Position Model Part 7
Appendix E
Unit Auxiliary Power
Consumption Simulink Model -
Process Model
158
Figure E.1: Unit Auxiliary Power Consumption Simulink Model - Process Model Part 1
159
Figure E.2: Unit Auxiliary Power Consumption Simulink Model - Process Model Part 2
160
Figure E.3: Unit Auxiliary Power Consumption Simulink Model - Process Model Part 3
161
Figure E.4: Unit Auxiliary Power Consumption Simulink Model - Process Model Part 4
162
Figure E.5: Unit Auxiliary Power Consumption Simulink Model - Process Model Part 5
163
Figure E.6: Unit Auxiliary Power Consumption Simulink Model - Process Model Part 6
164
Figure E.7: Unit Auxiliary Power Consumption Simulink Model - Process Model Part 7
Appendix F
Process Information Management
System (PIMS) Extracted Data
166
Figure F.1: Example of PIMS Data Extracted for the 7th of July 2015 (Siemens 2015)
Appendix G
Located Drawing Data
Below in table G.1 is a sample of the drawings collected during the data collection stage.
Drawing Number Drawing Title
A2-V-502446-01 Unit AC/DC Power Distribution: Single Line Diagram
A3-V-505763-01 Unit Single Line Diagram Typical for Four Units
A1-V-503115-01 Unit 6.6kV Board Switchgear Detail Diagram Tarong Power Station
A1-V-534312-036 Unit 4 A Boiler Feed Pump Suction Flow T4LAB11CF001 DCS Loop Diagram
A1-V-534312-040 Unit 4 B Boiler Feed Pump Suction Flow T4LAB12CF001 DCS Loop Diagram
A1-V-534312-044 Unit 4 C Boiler Feed Pump Suction Flow T4LAB13CF001 DCS Loop Diagram
A1-V-534311-115 Unit 4 A Forced Draft Fan Air Flow T4HLA01CF001 DCS Loop Diagram
Table G.1: A Sample of Drawings Located during the Data Gathering Stage (Hitachi 2003).
Appendix H
Energy Metering Manual Data
169
MB Address Cubicle Tag Type Serial C/T Ratio V/T Ratio Locn PLC1 #4A Unit 415 V Tx T0MET1030 KILO 59504 200/1 60/1 Unit 6.6kV MET22 #4A Precip Tx TOMET1031 KILO 59527 200/1 60/1 Unit 6.6kV MET23 #4A Unit 6.6 kV Tx T0MET1032 KILO 59528 2400/1 60/1 Unit 6.6kV MET24 #4 Bunkerinq Tx TOMET1033 KILO 59522 200/1 60/1 Unit 6.6kV MET25 #4C Unit 415 VTx T0MET1034 KILO 59530 200/1 60/1 Unit 6.6kV MET26 #4 Unit / Stn TIE T0MET1035 KILO 55149 2400/1 60/1 Unit 6.6kV MET27 #4B Unit 6.6 kV Tx T0MET1036 KILO 59523 2400/1 60/1 Unit 6.6kV MET28 #4B Unit 415 vTx T0MET1037 KILO 59525 200/1 60/1 Unit 6.6kV MET29 #4B Precip Tx T0MET1038 KILO 58529 300/1 60/1 Unit 6.6 V MET2
10 Fire Fighting 415 V Transf T0MET1001 KILO 59511 200/1 60/1 Stn 6.6kV MET111 B Ash Plant 415 V Transf T0MET1002 KILO 59514 200/1 60/1 Stn 6.6kV MET112 B Admin Area 415 V Tranf T0MET10O3 KILO 59519 200/1 60/1 Stn 6.6kV MET113 B1 Station 415 V Tranf T0MET1004 KILO 55156 200/1 60/1 Stn 6.6kV MET114 B Coal Recv 415 V Transf T0MET1005 KILO 59510 200/1 60/1 Stn 6.6kV MET115 D Ash Paint 415 v Transf T0MET1006 KILO 55151 200/1 60/1 Stn 6.6kV MET116 B2 Station 415 V Transf T0MET1007 KILO 59506 200/1 60/1 Stn 6.6kV MET117 B Cooling Tower 415 V Trf T0MET1008 KILO 59508 200/1 60/1 Stn 6.6kV MET118 B Station Transf T0MET1009 KILO 59537 2000/1 60/1 Stn 6.6kV MET119 C1 & C2 Station 415 V Transf T0MET1010 KILO 59512 200/1 60/1 Stn 6.6kV MET120 Gas Turbine Incomer T0MET1011 KILO 55152 2000/1 60/1 Stn 6.6kV MET121 Taron North Supply T0MET1012 KILO 59516 200/1 60/1 Stn 6.6kV MET122 A Station Transf T0 ET1013 KILO 59520 2000/1 60/1 Stn 6.6kV MET123 A Ash Plant 415 V Transf T0MET1014 KILO 59513 200/1 60/1 Stn 6.6kV MET124 A Admin Area 415 VTrf T0MET1015 KILO 59521 200/1 60/1 Stn 6.6kV MET125 A1 Station 415 V Transf T0MET1016 KILO 55155 200/1 60/1 Stn 6.6kV MET126 A Coal Rec 415 V transf T0MET1017 KILO 59517 200/1 60/1 Stn 6.6kV MET12728 CAsh Plant 415 V trf T0MET1018 KILO 55150 200/1 60/1 Stn 6.6kV MET129 A2 Station 415 v Trf T0MET1019 KILO 55147 200/1 60/1 Stn 6.6kV MET130 A Cooling Tower 415 V trf T0MET1020 KILO 59543 200/1 60/1 Stn 6.6kV MET131 CW Makeup 415 VTrf T0MET1021 KILO 59509 150/1 60/1 Stn 6.6kV MET132 #4A Unit Transf HV T0MET1039 KILO 59505 1600/1 60/1 X&Y PROT MET233 #4B Unit Transf HV T0MET1040 KILO 59505 1600/1 60/1 X&Y PROT MET234 #4 Exitation Trf HV T0MET1058 KILO 59526 200/1 60/1 X&Y PROT MET235 SPARE KILO 59518 200/1 60/1 Stn 6.6kV MET136 #4 Generator (Excitation Transf) T0MET1041 KILO 59531 12000/1 60/1 X&Y PROT MET237 #4A PA Fan T0 ET1042 EST-485 58803 150/1 60/1 Unit 6.6kV MET238 #4A FD Fan T0 ET1043 EST-485 58807 400/1 60 1 Unit 6.6kV MET 39 #4A ID Fan T0MET1044 EST-485 58802 500/1 60/1 Unit Q.6kV MET240 #4A Cond Extraction P/P T0MET1045 EST-485 58804 150/1 60/1 Unit 6.6kV MET241 #4A CW P/P T0MET1046 EST-485 58798 200/1 60/1 Unit 6.6kV MET242 #4A BFP T0MET1047 EST-485 58790 600/1 60/1 Unit 6.6kV MET243 #4C BFP T0MET1048 EST-485 58797 600/1 60/1 Unit 6.6kV MET244 #4B PA Fan T0MET1049 EST-485 58796 150/1 60/1 Unit 6.6kV MET245 #4B FD Fan T0MET1050 EST-485 58805 400/1 6Q)/1 Unit 6.6kV MET246 #4B ID Fan T0MET1051 EST-485 58791 500/1 60/1 Unit 6.6kV MET247 #4B Cond Extraction P/P T0MET1052 EST-485 58792 150/1 60/1 Unit 6.6kV MET248 #4B CW P/P T0MET1053 EST-485 58789 200/1 S0/1 Unit 6.6kV MET249 #4B BFP T0MET1054 EST-485 58799 600/1 60/1 Unit 6.6kV MET250 B Ash Water Reclaim T0MET1022 EST-485 58784 60/1 60/1 Stn 6.6kV MET151 B CW akeup Pump T0MET1023 EST-485 58800 75/1 §0/1 Stn 6.6kV MET152 R1 Conveyor Drive T0MET1024 EST-485 58793 75/1 ©0/1 Stn 6.6kV MET153 A CW Makeup Pump T0MET1025 EST-485 58787 75/1 60/1 Stn 6.6kV MET154 A Ash Water Makeup Pump T0MET1026 EST-485 58806 60/1 60/1 Stn 6.6kV MET155 #4A Treated Water Low Lift Pump T0MET1059 EST-485 58816 800/5 Not Reqd Unit 415V MET256 #4A Pulveriser T0MET1060 EST-485 58820 500/5 Not Reqd Unit 415V MET257 #4A Cond Polisher Booster Pmp T0MET1061 EST-485 58820 250/5 Not Fseqd Unit 415V ET258 #4C Pulveriser T0MET1062 EST-485 58822 500/5 Not Reqd Unit 415V MET259 #4Unit Pulv. A Sealing Air Fan T0MET1063 EST-485 58825 250/5 Not Reqd Unit 415V MET260 #4A Condenser Extraction pmp T0MET1064 EST-485 58827 200/5 60/1 Unit 6.6kV MET261 #4B Treated Water Low Lift Ppm T0MET1065 EST-485 51396 800/5 Not Reqd Unit 415V MET262 #4F Pulveriser T0MET1066 EST-485 51386 500/5 Not Reqd Unit 415V MET263 #4D Pulveriser T0MET1067 EST-485 58832 500/5 Not Reqd Unit 4i'5V MET264 #4B Cond Polisher Booster Pmp T0MET1068 EST-485 58833 250/5 Not Reqd Unit 415V MET265 #4Unit Pulv. B Sealing Air Fan T0MET1069 EST-485 58834 250/5 Not Reqd Unit 415V M5T266 #4B Condenser Air Extraction Pmp T0MET1070 EST-485 59122 200/5 Not Reqd Unit 415V MET267 #4E Pulveriser T0MET1071 EST-485 59124 500/5 Not Reqd Unit 415V MET268 #4B Pulveriser T0MET1072 EST-485 59123 500/5 Not Reqd Unit 415V MET269 #4Dearator Standby Pump T0MET1073 EST-485 58831 400/5 Not Reqd Unit 415V MET270 Pozzo Compressor No. 1 T0MET1027 KILO 59565 500/5 Not Reqd 1/2 Ash MET171 Pozzo Compressor No. 2 T0MET1028 KILO 59566 500/5 Not Reqd 1/2 Ash ME Cf72 Pozzo 3/4 Transfer Compressor T0MET1056 KILO 59567 500/5 Not Reqd 3/4 Ash MET273 Pozzo Compressor No. 3 T0MET1057 KILO 59568 600/5 Not Reqd 3/4 Ash MET274 Pozzo Compressor No. 4 T0MET1058 KILO 59569 600/5 Not Reqd 3/4 Ash MET2
1 Raw Water Pumps T0MET1075 KILO 87804 150/5 Not Reqd CT Area MET32 Pretreatment Plant T0MET1076 KILO 87803 150/5 Not Reqd CT Area MET33 Demin 'A' Board T0MET1077 KILO 87811 600/5 ot Reqd CT Area MET34 Demin 'B' Board T0MET1078 KILO 87805 600/5 Not Reqd CT Area MET35 A1 CRA Pump T0MET1079 KILO 48766 'lot Reqd Not Reqd CT Area MET36 B' CRA Pump T0MET1080 KILO 87808 'Jot Reqd ot Reqd CT Area MET37 Hydrogen Plant T0MET1081 KILO 87804 300/5 Not Reqd CT Area MET3
Figure H.1: List of Energy Meters Installed at Tarong Power Station (Godsmark 2001)
170
UNIT 4 6.6kV & 415V SWITCH ROOMUnit 4 'C 415V Switchboard Unit 6.6kV Switchboard containing 21 Metering Devices
Unit 4 'B' 15V Switchboard Unit 4 'A' 415 Switchboard
4A TREATED WATER LO LIFT PUMP CUBICLE ONLYCOM UNICATION LINE AMPLIFIER
To other panel Metering Devices
A A
120 OHMS >RX0+RXO-TXD+TXD-
ISOLATEDRS-422/488REPEATER
To Metering-t> Device
To supply CBTo neutral <8-
IPS-1A
1110
8
POWER SUPPLY 73 62 51 4
UNIT 4 GENERATOR & TRA SFORMERSPROTECTION ROOM
U4 T Protection Panel U4 'X' Protection Panel
UNITS 3 & 4 STATION RELAY ROOM
RS232/RS485INTERFACE
rdScreen Bbd<
From ModiconModicon PLC PlugNODE 25 et 2
P r Supply
j ft3/A P Scv%i£ iS P"
UNITS 3 & 4 ASH & DUST S ITCH ROOM. scSSI?
Unit 3&4 ASH & DUST 415V Switchboard
L Pozzolanic Feeder Metering
#
rsi A
UNIT 4 E ERGY METERING CABLING & ARRANGEMENTSSh.12 of 12
Figure H.2: Drawing of Typical Energy Meter Installation at Tarong Power Station
(Godsmark 2001)
171
ENERGY2 Msg69_Data_7 REALENERGY2 Msg69_Data_9 REALENERGY2 Msg70_Data_1 REALENERGY2 Msg70_Data_3 REALENERGY2 Msg70_Data_5 REALENERGY2 Msg70_Data_7 REALENERGY2 Msg70_Data_9 REALENERGY2 Msg70_Data_11 REALENERGY2 Msg71_Data_1 REALENERGY2 Msg71_Data_3 REALENERGY2 Msg71_Data_5 REALENERGY2 Msg71_Data_7 REALE ERGY2 Msg72lbata_1 EAL-
ENERGY2 Msg72_Data_3 REALENERGY2 Msg72_Data_5 REALENERGY2 Msg72_Data_7 REALENERGY2 Msg72_Data_9 REALENERGY2 Msg72_Data_11 REALENERGY2 Msg73_Data_1 REALENERGY2 Msg73_Data_3 REALENERGY2 Msg73_Data_5 REALENERGY2 Msg73_Data_7 REALENERGY2 Msg73_Data_9 REALENERGY2 Msg74_Data_1 REALENERGY2 Msg74_Data_3 REALENERGY2 Msg74_Data_5 REALENERGY2 Msg74_Data_7 REALjENERGY2 Msg74_Data_9 REALENERGY2 Msg74_Data_11 REALENERGY2 Msg75_Data_1 REALENERGY2 Msg75_Data_3 REALENERGY2 Msg75_Data_5 REALENERGY2 Msg75_Data_7 REALENERGY2 Msg r6_Data~T ~ TftAL ENERGY2 Msg76_Data_3 REALENERGY2 Msg76_Data_5 REALENERGY2 Msg76_Data_7 REALENERGY2 Msg76_Data_9 REALENERGY2 Msg76_Data_11 REALENERGY2 Msg77_Data_1 REALENERGY2 Msg77_Data_3 REALENERGY2 Msg77_Data_5 REALENERGY2 Msg77_Data_7 REALENERGY2 Msg77_Data_9 REALENERGY2 Msg78_Data_1 REALENERGY2 Msg78_Data_3 REALENERGY2 Msg78_Data_5 REALENERGY2 Msg78_Data_7 REALENERGY2 Msg78_Data_9 REALENERGY2 Msg78_Data_11 REALENERGY2 Msg79_Data_1 REALENERGY2 Msg79_Data_3 REALENERGY2 Msg79_Data_5 REALENERGY2 Msg79_Data_7 REALEfTERGY? Msg80_Data_1 ' E L
,iNERGY2 Msg80_Data_3 REALENERGY2 Msg80_Data_5 REALENERGY2 Msg80_Data_7 REALENERGY2 Msg80_Data_9 REALENERGY2 Msg80_Data_11 REALENERGY2 Msg81_Data_1 REALENERGY2 Msg81 J0ata_3 REALENERGY2 Msg81_Data_5 REALENERGY2 Msg81_Data_7 REALENERGY2 Msg81_Data_9 REALENERGY2 Msg82_Data_1 REALENERGY2 Msg82_Data_3 REALENERGY2 Msg82_Data_5 REALENERGY2 Msg82_Data_7 REALENERGY2 Msg82_Data_9 REALENERGY2 Msg82_Data_11 REALENERGY2 Msg83_Data_1 REALENERGY2 Msg83_Data_3 REALENERGY2 Msg83_Data_5 REALENERGY2 Msg83_Data_7 REAL
" R'G?2 Msg84_Data_1 RIALENERGY2 Msg84_Data_3 REALENERGY2 Msg84_Data_5 REALENERGY2 Msg84_Data_7 REALENERGY2 Msg84_Data_9 REALENERGY2 Msg84_Data_11 REALENERGY2 Msg85_Data_1 REALENERGY2 Msg85_Data_3 REALENERGY2 Msg85_Data_5 REALENERGY2 Msg85_Data_7 REAL
404726 4A PA Fan Not Used404728 4A PA Fan Total VARs404730 4A PA Fan A Ph Volts404732 4A PA Fan B Ph Volts404734 4A PA Fan C Ph Volts404736 4A PA Fan A Ph Amps404738 4A PA Fan B Ph Amps404740 4A PA Fan C Ph Amps404742 4A PA Fan A Ph Watts404744 4A PA Fan B Ph Watts404746 4A PA Fan C Ph Watts404748 4A PA Fan Frequency
'404750 A PD Fan 3Ph olts" 404752 4A FD Fan 3Ph Amps404754 4A FD Fan 3Ph Watts404756 4A FD Fan 3Ph VARs404758 4A FD Fan 3Ph VAs404760 4A FD Fan 3Ph PF404762 4A FD Fan Watts MD404764 4A FD Fan VA MD404766 4A FD Fan Total Watts404768 4A FD Fan Not Used404770 4A FD Fan Total VARs404772 4A FD Fan A Ph Volts404774 4A FD Fan B Ph Volts404776 4A FD Fan C Ph Volts404778 4A FD Fan A Ph Amps404780 4A FD Fan B Ph Amps404782 4A FD Fan C Ph Amps404784 4A FD Fan A Ph Watts404786 4A FD Fan B Ph Watts404788 4A FD Fan C Ph Watts404790 4A FD Fan Frequency
¦?OT7g2 4A'IU ah'3P 'V6irs 404794 4A ID Fan 3Ph Amps404796 4A ID Fan 3Ph Watts404798 4A ID Fan 3Ph VARs404800 4A ID Fan 3Ph VAs404802 4A ID Fan 3Ph PF404804 4A ID Fan Watts D404806 4A ID Fan VA MD404808 4A ID Fan Total Watts404810 4A ID Fan Not Used404812 4A ID Fan Total VARs404814 4A ID Fan A Ph Volts404816 4A ID Fan BPh Volts404818 4A ID Fan C Ph Volts404820 4A ID Fan A Ph Amps404822 4A ID Fan B Ph Amps404824 4A ID Fan C Ph Amps404826 4A ID Fan A Ph Watts404828 4A ID Fan B Ph Watts404830 4A ID Fan C Ph Watts404832 4A ID Fan Frequency
NOT USED
A
NOT USED
jo ti
NOT USED
(°i-
404836 4A Cond Extr Pmp 3Ph Amps404838 4A Cond Extr Pmp 3Ph Watts404840 4A Cond Extr Pmp 3Ph VARs404842 4A Cond Extr Pmp 3Ph VAs404844 4A Cond Extr Pmp 3Ph PF404846 4A Cond Extr Pmp Watts MD404848 4A Cond Extr Pmp VA MD404850 4A Cond Extr Pmp Total Watts404852 4A Cond Extr Pmp Not Used404854 4A Cond Extr Pmp Total VARs404856 4A Cond Extr Pmp A Ph Volts404858 4A Cond Extr Pmp B Ph Volts404860 4A Cond Extr Pmp C Ph Volts404862 4A Cond Extr Pmp A Ph Amps404864 4A Cond Extr Pmp B Ph Amps404866 4A Cond Extr Pmp C Ph Amps404868 4A Cond Extr Pmp A Ph Watts404870 4A Cond Extr Pmp B Ph Watts404872 4A Cond Extr Pmp C Ph Watts404874 4A Cond Extr Pmp Frequency" OTSTBAa CWPmp~3PhVolts404878 4A CW Pmp 3Ph Amps404880 4A CW Pmp 3Ph Watts404882 4A CW Pmp 3Ph VARs404884 4A CW Pmp 3Ph VAs404886 4A CW Pmp 3Ph PF404888 4A CW Pmp Watts MD404890 4A CW Pmp VA MD404892 4A CW Pmp Total Watts404894 4A CW Pmp Not Used
( -TB P
NOT USED
I I V
I of
NOT USED
Figure H.3: Example of Typical Energy Meter PLC Registers and Functions (Godsmark 2001)
Appendix I
Stanwell Corporation Energy
Reductions Recommendations
Report
173
Figure I.1: Part 1 of Energy Reduction Recommendations Memorandum to David Janes
174
Figure I.2: Part 2 of Energy Reduction Recommendations Memorandum to David Janes